Summary
granting preliminary injunction to plaintiff software developer where defendant distributor and licensee end users were bound by confidentiality provisions in both the Distribution Agreement and the End User License Agreements
Summary of this case from Broker Genius, Inc. v. ZaltaOpinion
CIVIL ACTION NO. 4:13-CV-3182
12-10-2013
MOBIUS MEDICAL SYSTEMS, LP, Plaintiff, v. SUN NUCLEAR CORPORATION, Defendant.
OPINION AND ORDER
Pending before the Court is Plaintiff Mobius Medical Systems, L.P.'s ("Mobius") Motion for Preliminary Injunction (Doc. 7). Upon review and consideration of the motion, response, reply, the evidence submitted, and the testimony presented at the hearing on November 25 and 26, 2013, the Court concludes that Mobius's motion should be granted.
I. Background
On October 11, 2013, Mobius filed the instant suit in the 129th District Court in Harris County, Texas against Defendant Sun Nuclear Corporation ("Sun Nuclear") alleging breach of contract, trade secret misappropriation, trade dress infringement, and tortious interference with existing and prospective contracts. Pl.'s Original Pet. (Doc. 1, Ex. 2). Sun Nuclear removed the suit to this Court on October 29, 2013 under the Court's diversity jurisdiction. Def.'s Notice of Removal (Doc. 1). On November 7, 2013, Mobius filed the instant Motion for Preliminary Injunction (Doc. 7) based on its claim for trade secret misappropriation. Sun Nuclear has also filed motions to Stay Litigation Pending Arbitration (Doc. 16) and Dismiss pursuant to Rule 12(b)(6) (Doc. 17).
A. The Parties
Mobius was founded in 2010 by Dr. Nathan Childress, a graduate of M.D. Anderson's medical physics program. Aff. of Nathan Childress ¶ 2-3 (Doc. 7, Ex. 1). Mobius is a developer of software used in the field of radiation oncology. Doc. 7, Ex. 1 ¶ 3. Mobius's most commercially successful software product is DoseLab. Doc. 7, Ex. 1 ¶ 4. DoseLab provides quality assurance (QA) for all types of radiation oncology linear accelerators used in the treatment and imaging of cancer patients. Doc. 7, Ex. 1 ¶ 4. QA enables those administering radiation treatments to ensure that the machine is operating properly and that the correct dosage is being delivered to the patient. Doc. 7, Ex. 1 ¶ 5. The industry standards for the QA of medical linear accelerators are outlined in a report published by the American Association of Physicists in Medicine known as the Task Group 142 Report ("TG-142 Report"). See TG-142 Report, Doc. 7, Ex. 2 at 4197-4198. "Nearly all healthcare institutions and treatment centers in the U.S. and abroad look to the TG-142 Report guidelines to ensure that their radiation treatment and image guidance machines are working properly. The TG-142 Report is the de facto standard for medical linear accelerator QA." Doc. 7, Ex. 1 ¶ 6.
According to Mobius's website, the development history of DoseLab is as follows:
"DoseLab began as a clinical and research tool for U.T. MD Anderson Cancer Center. It was written primarily by Nathan Childress as part of his Ph.D. dissertation on film dosimetry, under the supervision of his advisor Dr. Isaac Rosen. Additional coding was performed by Dr. Isaac Rosen and Terry Vantreese. It was designed to quickly perform IMRT QA comparisons while retaining the data for future analysis. It was distributed worldwide under the MIT Open Source license, and has been downloaded over 5,000 times. Its original source code is copyright Nathan Childress and U.T. MD Anderson Cancer [Center]. The original version can still be found at Sourceforge.net.
As DoseLab's popularity grew, its maintenance and development became more than a full-time job. Dr. Childress performed a complete overhaul of DoseLab by adding dozens of features, upgrading each image analysis tool and processing routine to be more robust, improving overall stability, enhancing the clinical
workflow, and assembling a team of experts to provide customer support. In addition to modifying more than 40% of DoseLab's source code, many new features were added. The end result is DoseLab Pro: a fast, simple, and powerful software system that can perform everything from patient-specific SBRT QAs to advanced research analyses using automatically generated clinical data." See Mobius Website: DoseLab History (Doc. 14, Ex. 3).
Sun Nuclear is a privately held corporation with "over 200 employees, including 16 physicists and 12 software developers." Decl. of Matthew Daniels ¶ 4 (Doc. 14, Ex. 1). Since Sun Nuclear was founded in 1984 it has become a world leader in the "development manufacture, and sale of radiation measurement instrumentation and software." Doc. 14, Ex. 1 ¶ 4. In 2010, Sun Nuclear's president, William Simon, approached Dr. Childress with a proposal that Mobius and Sun Nuclear enter into a distribution agreement for the sale of DoseLab. Doc. 1, Ex. 2 ¶ 15. At the time, Sun Nuclear did not have a software product that performed TG-142 analysis. Doc. 7, Ex. 1 ¶ 9. Doc. 7 at 3. Because of Sun Nuclear's strong reputation in the radiation oncology business, and its existing sales and distribution infrastructure, Mobius negotiated a release from its original distributor and signed an agreement for Sun Nuclear to become the exclusive distributor of DoseLab software. Doc. 7, Ex. 1 ¶ 10.
B. The Agreements
In March 2011, DoseLab entered into an exclusive software distribution agreement ("the Distribution Agreement") with Sun Nuclear. Doc. 7, Ex. 3. Under the terms of the Distribution Agreement, Sun Nuclear agreed to use its best efforts to promote and sell DoseLab in exchange for a percentage of the sales revenue. Doc. 7, Ex. 3, ¶ 1.3. Under section 7 of the Distribution Agreement, Sun Nuclear agreed that it would not sell or promote any products that were directly competitive with DoseLab during the life of the Distribution Agreement and for six months after. Doc. 7, Ex. 3, ¶ 7.
Section 7 of the Distribution Agreement provided: 7. Distributor covenants and agrees that it (a) will not serve as a distributor, dealer, or sales agent of, and not to sell, license, lease or market, any third party products which are Directly Competitive with the Products, at any time during the term of this Agreement and for a period of six (6) month after the expiration or earlier termination; and (b) will not to sell any of its own products which are Directly Competitive with the Products at any time during the term of this Agreement and for a period of six (6) months thereafter. For purposes of the foregoing covenant, any and all products whose primary function is film or EPID based image analysis software for radiation oncology linear accelerator QA as defined in AAPM TG-142, and treatment log analysis software whose primary function is per fraction QA, will be deemed to be "Directly Competitive" with the Products.
Under section 9 of the Distribution Agreement, Sun Nuclear agreed that during the term of the agreement and "at all times thereafter," Sun Nuclear would not use or disclose Mobius's confidential information. Sections 9.1-9.2 provide:
9.1 Each party hereby acknowledges that it may be exposed to confidential and proprietary information belonging to or supplied by the other party or relating to its business operations and affairs including, without limitation, customer lists and customer opportunities, market intelligence, pricing, market share, revenue, discount and IP knowledge and other technical information (including any Functional Design, Technical Design, drawings, analysis, research, processes, computer programs, methods, ideas, "know how" and the like) business information (sales and marketing research, materials, plans, accounting, and financial information, personnel records and the like) and other information designated as confidential expressly or by the circumstances in which it is provided ("Confidential Information"). Confidential Information does not include (i) information already known or independently developed by the recipient outside the scope of this project by personnel not having access to Confidential Information; (ii) information in the public domain through no wrongful act of the recipient, or (iii) information received by the recipient from a third party who was free to disclose it.Doc. 7, Ex. 3, ¶¶ 9.1-9.2.
9.2 With respect to each party's Confidential Information, and except and expressly authorized herein, the each [sic] party hereby agrees that during the Term hereof and at all times thereafter it shall not use or disclose such Confidential Information to any person or entity, except to its own employees having a "need to know" (and who are themselves bound by similar nondisclosure restrictions), and to such other recipients as the other party may approve in writing; provided, that all such recipients shall have first executed a confidentiality agreement in a form acceptable to the owner of such information. [Sun Nuclear] may not: (i) alter or remove from any Product or associated documentation owned or provided by [Mobius] any proprietary, copyright, trademark or trade secret legend, or (ii) attempt to decompile, disassemble or reverse engineer [Mobius]'s Products (and any information derived in violation of such covenant shall automatically be deemed Confidential Information owned
exclusively by [Mobius]). Each party shall use at least the same degree of care in safeguarding the other party's Confidential Information as it uses in safeguarding its own confidential information."
Under Section 13.1, either party had a right to terminate the agreement: "by giving one hundred and eighty (180) days written notice of termination to the other party." Doc. 7, Ex. 3 ¶ 13.1. Under Section 15.3, the parties agreed that both Sections 7 and 9 would "survive the expiration or earlier termination of th[e] Agreement for any reason." Doc. 7, Ex. 3 ¶ 15.3.
By all accounts, Sun Nuclear's sales of DoseLab exceeded expectations. Doc. 7, Ex. 1 ¶ 11; Doc. 14, Ex. 1 ¶ 7. However, on March 8, 2013, Sun Nuclear CEO Jeff Simon informed Mobius of his intent to terminate the Distribution Agreement. Doc. 7, Ex. 1 ¶ 13. Sun Nuclear had decided to develop its own proprietary software for TG-142 QA testing. Doc. 14, Ex. 1 ¶ 9. Sun Nuclear exercised its rights under the termination provision of the Distribution Agreement, and on April 10, 2013, Sun Nuclear and Mobius entered into the Transition Agreement to govern the termination of Sun Nuclear's exclusive distributorship. See Doc. 7, Ex. 4. Under the Transition Agreement, Sun Nuclear retained exclusive rights to distribute DoseLab in the United States until September 8, 2013, and exclusive rights to distribute DoseLab internationally until December 31, 2013. Doc. 7, Ex. 4 ¶ 1(a). In addition, Sun Nuclear could sell a product competitive to DoseLab as early as September 8, 2013. Doc. 7, Ex. 4 ¶ 2. This provision reflected a significant concession on the part of Mobius, as the Distribution Agreement prohibited competition for a period of six months after expiration or termination of that Agreement. Doc. 7, Ex. 3, ¶ 7. Section 4 of the Transition Agreement stated that a number of provisions from the Distribution Agreement, including the Non-Compete clause in Section 7, would no longer be applicable in their entirety. Doc. 7, Ex. 4 ¶ 4. Those provisions that were not specifically referenced in Section 4 of the Transition Agreement, including the Covenant Not to Use or Disclose in Section 9, remained in effect. Doc. 7, Ex. 4 ¶ 5.
C. Post-Termination of the Distributorship
On September 10, 2013, Sun Nuclear released its own TG-142 solution software product: ImagePro. Doc. 7 at 8. In the weeks and months between the termination of the Distribution Agreement and the release of ImagePro, Sun Nuclear's behavior raised Dr. Childress's suspicions that Sun Nuclear was attempting to reverse engineer DoseLab to create its own product. Doc. 7, Ex. 1 ¶ 27. From its webserver logs, Mobius detected that an IP address at Sun Nuclear's headquarters in Brevard County, Florida accessed the DoseLab software approximately 750 times. Doc. 7, Ex. 1 ¶ 27. According to Mobius, this demonstrates that while Sun Nuclear was developing ImagePro, Sun Nuclear's engineers "repeatedly launched DoseLab, decompiled, disassembled, or otherwise reverse engineered DoseLab's programming and improperly used DoseLab's features as a guide for ImagePro's development." Doc. 7 at 17-18. Sun Nuclear attributes this disproportionate webserver access to the fact that Sun Nuclear is still the exclusive distributor of DoseLab and Sun Nuclear employees are still responsible for familiarizing themselves with the software updates and providing customer support. Doc. 14 at 15-16.
For purposes of comparison, Mobius logged less than ten messages per individual DoseLab trial user during the same period, and two Houston-area treatment facilities with a combined total of four active linear accelerators sent only 92 messages during the same 9-month period. Doc. 7, Ex. 1 ¶ 27.
Mobius claims that in the weeks since the release of ImagePro, Sun Nuclear has "wrongly informed potential customers that it was no longer a distributor of DoseLab, improperly tried to get customers to purchase ImagePro instead," attempted to market ImagePro as an "upgrade" of DoseLab to existing DoseLab licensees, and "purposefully exploited consumer confusion" between DoseLab and ImagePro. Doc. 7, Ex. 1 ¶ 31-35. In late October, after Mobius filed the instant suit, Sun Nuclear began "offering an ImagePro 'swap' to licensees with pending DoseLab purchase orders." Doc. 7, Ex. 1 ¶ 39. Mobius claims these and other actions by Sun Nuclear have "deprived Mobius of DoseLab sales and opportunities, tarnished the goodwill DoseLab has fostered among its licensees, and diminished the value of Mobius's intellectual property." Doc. 7 at 26.
D. Trade Secrets
As a whole, the overall operations and functionality of DoseLab and ImagePro are obviously similar. For example, in order to perform the monthly QA required in the TG-142, DoseLab requires a physicist to acquire 15 images using five phantoms. Doc. 7, Ex. 1 ¶ 28. PIPSpro, the closest competing product to DoseLab requires that the physicist acquire 26 images using eight phantoms. Doc. 7, Ex. 1 ¶ 28. Like DoseLab, ImagePro requires 15 images using five phantoms. Even to the Court's untrained eye, DoseLab and ImagePro share other similarities in terms of the overall look and feel of the software programs to a degree not shared between DoseLab and other competing products including PIPSpro, Varian, and Film QA Pro. Mobius attributes the similarities between the products to trade secret theft. Sun Nuclear attributes the similarities to the fact that both programs are engaged in analyzing the same phantoms and using the same standards outlined in the TG-142.
In its motion for preliminary injunction, Mobius states that its "claim for trade secret misappropriation stems from [Sun Nuclear]'s dissembling, decompiling, or reverse engineering DoseLab's algorithms, visualization techniques, configurations and parameters, and other DoseLab-exclusive technology in violations of [Sun Nuclear]'s express covenant." Doc. 7 at 21. At the hearing, Dr. Childress admitted that he had not used the ImagePro software, nor had he seen its source code. Mobius bases its claim for trade secret misappropriation on the "striking functional, parametric, and visual similarity between DoseLab and ImagePro" as it appears in the ImagePro marketing materials. Doc. 7 at 8. Sun Nuclear contends that similarities in the screenshots of DoseLab and ImagePro while running similar functions is not evidence that the underlying code or algorithms were copied. Doc. 14 at 13. Mobius counters that the screenshots simply evidence that the "parameters and results of Mobius's proprietary algorithms and formulas" were copied. Doc. 22 at 8.
Sun Nuclear insists that DoseLab was not used as a blueprint or roadmap for the development of ImagePro. Doc. 14 at 6. Sun Nuclear points out that it was never provided anything more than a commercially available version of DoseLab and it never had access to DoseLab's source code, algorithms, or other non-public technical information. Doc. 14, Ex. 1 ¶ 7. In its response to Mobius's motion, Sun Nuclear emphasizes the differences between DoseLab and ImagePro. Unlike DoseLab, which is installed on an end-user's PC, ImagePro is a browser-based application which runs on an integrated data device. Doc. 14, Ex. 1 ¶ 10. ImagePro was written using HTML and JavaScript for the user interface and C# and C for the backend, while DoseLab uses MATLAB programming tools. Doc. 14, Ex. 1 ¶ 10. Sun Nuclear claims that ImagePro was developed by a team of 30 Sun Nuclear employees contributing over 18,889 hours of work. Doc. 14, Ex. 1 ¶ 11. "Presumably," it states, "it would not have required a 30 person team more than 18,000 hours of effort to simply 'knock-off' DoseLab." Doc. 14 at 8. Under cross-examination, Dr. Matthew Daniels, Sun Nuclear's project manager for the team that developed the ImagePro software, testified that 5,000 hours were spent developing the "major TG-142 functionalities," and the remaining 13,000-14,000 hours were spent on other miscellaneous tasks. ImagePro's user interface was purchased from Bootstrap, a "front-end toolkit for rapidly developing web applications" offered through Twitter. Mark Otto, Bootstrap from Twitter, https://dev.twitter.com/blog/bootstrap-twitter (August, 19, 2011).
Mobius claims that Sun Nuclear has misappropriated DoseLab's trade secrets including its formulas, algorithms, reference images, example images, and configurations files. To illustrate how these trade secrets were misappropriated, Mobius presented testimony and evidence on the following features of DoseLab which it says that ImagePro copied: kV/MV imaging QA, multi-leaf collimator QA, MLC strip test, cone beam CT imaging, and starshot imaging. Sun Nuclear claims that many of these features are outlined in the TG-142 Report, "and their display is common to all similar software products that include TG-142 functionality." Doc. 14 at 9. In addition, Sun Nuclear points out that some of these features are merely "aesthetic elements and data displayed as part of DoseLab's user interface." Doc. 14 at 9. Since a program's user interface is publically visible, it cannot possibly constitute a trade secret. Doc. 14 at 9. According to Sun Nuclear, both the TG-142 Report and other third-party products which pre-date DoseLab provide these same features. Doc. 14 at 12.
1. kV/MV Imaging QA
"Imaging systems used with medical linear accelerators use kilovoltage (kV) and megavoltage (MV) planar imaging to position patients and ensure radiation is accurately delivered." See TG-142 Report, Doc. 7, Ex. 2 at 4207-4208. QA of kV/MV imaging systems often utilizes "phantoms," described as "specially designed objects used to evaluate the performance of an imaging device by acting as a surrogate for human tissue or organs." Doc. 14, Ex. 1 ¶ 13. In DoseLab, Mobius invented a new algorithm to perform kV/MV imaging QA analysis of multiple types of phantoms from different manufacturers, also known as "multi-phantom analysis." Doc. 7 at 8. This feature is one of DoseLab's strongest selling points. Doc. 7, Ex. 1 ¶ 18. In its unique and exclusive algorithm, DoseLab "uses reference images to overlay red numbered boxes representing various regions of interest (ROIs)." No other TG-142 solution software product uses reference images except for DoseLab, and now ImagePro. DoseLab also generates kV/MV imaging results criteria that is unique in the industry including "scaling discrepancy, minimum uniformity, X positioning offset, and Y positioning offset." Doc. 7, Ex. 1 ¶ 19. Dr. Childress testified that he believes Sun Nuclear has misappropriated DoseLab's kV/MV imaging formulas, algorithms, reference images, and configuration files. According to Mobius, ImagePro copies DoseLab's results criteria verbatim. Doc. 7 at 9. Also, like DoseLab, ImagePro overlays red numbered boxes with ROIs with the same numbering scheme, same colors, and same positioning as DoseLab . . . ImagePro even copies DoseLab's internal configuration parameters and algorithm designed to analyze multiple phantoms from multiple manufactures. ImagePro also uses ROI coordinates identical to those used by DoseLab to at least three significant digits. Doc. 7, Ex. 1 ¶ 19.
Sun Nuclear argues that it did not copy DoseLab's kV/MV imaging QA. It argues that the use of red boxes and a similar numbering scheme and positioning are not trade secrets and that these characteristics do not reveal anything about the algorithms underlying ImagePro. Doc. 14 at 17. In support of its argument, Sun Nuclear points to another competing product, PIPSpro Version 4.2, which uses a similar "stoplight" color scheme. See PIPSpro Version 4.2 Manual (Doc. 14, Ex. 6 at 108). In addition, Sun Nuclear argues that using numbers 1 through 9 to identify points on a phantom is not a trade secret, nor are the coordinate areas on a phantom. Doc. 14 at 18. Likewise, Sun Nuclear claims that the particular results criteria chosen by ImagePro is known both from the TG-142 Report, the public display of DoseLab, and common knowledge within the industry, and therefore cannot constitute a trade secret. Doc. 14 at 19. Lastly, Sun Nuclear claims that DoseLab and ImagePro yielded different kV and MV results when processing the same image as conclusive evidence that the programs rely on different algorithms. Doc. 14, Ex. 1 ¶ 12. Dr. Childress responded to this argument at the hearing by stating that the similarities of the results (most within .01) reveal that the algorithms underlying DoseLab and ImagePro are virtually identical.
Sun Nuclear claims that multi-phantom kV/MV analysis became standard in the industry well before the release of ImagePro; however, it is undisputed that at the time that Sun Nuclear and Mobius entered into the Distribution Agreement, DoseLab was the only imaging QA program that had multi-phantom analysis capabilities. Doc. 7, Ex. 1 ¶ 18; Doc. 14, Ex. 1 ¶ 14. kV/MV functionality is addressed in the TG-142 Report. See TG-142 Report. However, Mobius's expert witness, Dr. Stephen F. Kry, testified that the TG-142 Report does not define ROIs, coordinates for ROIs, or baselines for kV/MV imaging. Nor does it describe procedures for kV/MV imaging using a reference image, or even procedures replicating the TG-142 standard. The ultimate responsibility for defining these criteria falls on the medical physicist.
2. Multi-Leaf Collimator QA
"A multi-leaf collimator (MLC) is a device made of thick metal leaves that move to dynamically shape a radiation treatment beam. Every leaf is monitored 20-100 times per second. A recording of the leaf position is called an MLC Log. MLC QA is essential to all clinics that use MLCs because MLC performances defines how precisely radiation is delivered to a patient." Doc. 7, Ex. 1 ¶ 20. The TG-142 Report addresses the parameters of MLC testing. See TG-142 Report, Doc. 7, Ex. 2 at Table V.
For DoseLab MLC QA, Mobius designed an "innovative graphical leaf representation using red, yellow, and green rectangles and informative mouse-over tooltips to reflect the results of the MLC log." Doc. 7, Ex. 1 ¶ 21. DoseLab's "results, configuration parameters, and method for visualizing and displaying those results are found in no other product—except now ImagePro." Doc. 7, Ex. 1 ¶ 21. According to Mobius, ImagePro "directly copies DoseLab's graphical leaf representation, uses the same red, yellow, and green display parameters, the same mouse-over tooltip text, and even copies DoseLab's unique display of multiple individual file results ordered in a table." Doc. 7 at 11.
Sun Nuclear claims that Mobius was not the first or only TG-142 solution software developer to display MLC results visually, and even if it were, Mobius would not be entitled to trade secret protection of a visual concept appearing in the user interface of the commercially available version of DoseLab. Doc. 14 at 21. In addition, Sun Nuclear argues that the use of "stoplight" colors to depict acceptable and problematic results is common in the industry. Doc. 14 at 22. With regard to the mouse-over tooltips that appear in ImagePro, Sun Nuclear states that such data displays are "ubiquitous in the industry and in MLC logs specifically." Doc. 14 at 22 (citing Aarhus software MLC Log (Doc. 14, Ex. 12); Doc. 14, Ex. 1 ¶ 21.).
3. MLC Strip Test
"Another method of MLC QA is the MLC strip test, which the TG-142 Report recommends be done weekly. During an MLC strip test, an image is taken by aligning the MLC's leaves to form a series of strips. The image is analyzed to determine each leaf's deviation from center. DoseLab's method of displaying [a] histogram representing differences of the individual leaves is also unique to DoseLab." Doc. 7, Ex. 1 ¶ 22. According to Mobius, ImagePro directly copies DoseLab's strip test method of displaying individual leaf location. Doc. 7 at 12.
Sun Nuclear argues that because the MLC strip test's graphical depictions are displayed in the user interface of DoseLab it cannot be a trade secret. Doc. 14 at 23. Also, Sun Nuclear points to three different third party products that use vertical lines, horizontal lines, histograms, or some combination thereof, to display MLC strip test results as evidence that such features are common in the industry. Doc. 14 at 23-24 (citing RIT MLC screenshot (Doc. 14, Ex. 13); RIT Leaf Representations screenshot (Doc. 14, Ex. 14); Varian Dynalog histogram screenshot (Doc. 14, Ex. 15); FilmQA Pro Leaf screenshot (Doc. 14, Ex. 16)).
4. Cone Beam CT Imaging
"Cone beam CT imaging provides a wealth of data regarding the proper alignment of a patient to a treatment beam. Accurately positioning patients based on their internal anatomy (bones, organs, etc.) is critical to safe and effective treatment. Thus, most new treatment machines include integrated CT imaging capabilities. Like kV/MV imaging, CT imaging QA uses phantoms, but a CT image consists of 5-200 individual images taken of different slices in a phantom." Doc. 7, Ex. 1 ¶ 23. In order to display these results, "DoseLab uses a distinct numbering scheme, distinct number of ROIs, distinct color-schemes and shapes for identifying certain ROIs, and distinct results parameters that correspond to the scaling discrepancy, minimum uniformity, and geometric distortion." Doc. 7, Ex. 1 ¶ 24. According to Mobius, DoseLab's method for displaying cone beam CT imaging is entirely unique, and ImagePro directly copies every aspect of that method. Doc. 7 at 13. Mobius states that "ImagePro uses the same results parameters, and employs the same exclusive ROI numbering scheme, positioning, and unique shape . . . ImagePro copies the exact look and feel of each display module." Doc. 7 at 14.
Sun Nuclear claims that it did not copy any of Mobius's cone beam CT trade secrets, and that even if it had, Mobius cannot claim trade secret protection for its numbers and coordinates which are visible from DoseLab's user interface. Doc. 14 at 25. According to Sun Nuclear, the key functions of ImagePro's cone beam CT imaging, including compatibility with the two phantoms specified for this type of imaging and the display of parameters set forth in the TG-142 Report (geometric distortion, spatial resolution, contrast, HU constancy, uniformity, and noise), are standard in the industry for any QA program. Doc. 14 at 25. Sun Nuclear points to a third party product which also analyzes the same phantom and parameters for cone beam CT imaging. Doc. 14 at 26 (citing PIPSpro datasheet (Doc. 14, Ex. 7)). It claims that DoseLab's "exclusive numbering scheme" is nothing more than a numerical order, and its coordinates simply correspond to the critical areas of phantoms.
Dr. Daniels testified at the hearing regarding each of the various software features at issue in Mobius's motion. During Dr. Daniels's cross-examination, he was asked why he and his team had chosen ROIs "2" and "9" for ImagePro's cone beam CT coordinates when Catphan, the developer of the cone beam CT phantom, had not specifically identified those ROIs for analysis. Dr. Daniels had no explanation for Sun Nuclear's decision with regard to those particular ROIs other than to say that they were the ones chosen by DoseLab, and ImagePro was created to compete with DoseLab.
5. Starshot
"A starshot image is taken by delivering a very narrow strip of radiation, rotating a component of the delivery system, and taking a new image. The sum of these images creates a starshot pattern." Doc. 7, Ex. 1 ¶ 25. DoseLab created an exclusive example file image for the starshot. Doc. 7 at 16. Mobius claims that Sun Nuclear used the same example file to create a similar starshot image which it has used as a cornerstone of ImagePro's marketing. Doc. 7 at 17.
Sun Nuclear insists that it did not use any of DoseLab's example file images, algorithms, or configuration files to create its starshot image. Doc. 14 at 30. To highlight the differences between the starshot analyses of DoseLab and ImagePro, Sun Nuclear states that the images are "fundamentally different in appearance than those of DoseLab ...in coloring, width and strip grouping." Doc. 14 at 30. In addition, Sun Nuclear claims a comparison test of the two programs revealed that "ImagePro can open and process 9 files that DoseLab cannot, while there are 3 files that DoseLab can open and process that ImagePro cannot." Doc. 14, Ex. 1 ¶ 12. The results of this comparison test, "conclusively establish that the algorithms underlying the respective program's starshot functionality are not the same." Doc. 14 at 31. Sun Nuclear also argues that the use of starshot images in QA analysis predates DoseLab, and points to PIPSpro version 4.2, released in 2008, which includes starshot images. Doc. 14 at 29 (citing PIPSpro Version 4.2 Manual (Doc. 14, Ex. 6 at 22-23)). Sun Nuclear also points out that the open source version of DoseLab, for which the source code was made public, included starshot analysis. Doc. 14, Ex. 1 ¶ 29. Accordingly, "neither the use of starshot images in DoseLab or the underlying algorithms, can be considered trade secrets." Doc. 14 at 30.
II. Mobius's Motion for Preliminary Injunction
Mobius argues that ImagePro is "nothing more than a clone of DoseLab created through indiscriminate copying of DoseLab's technology." Doc. 7 at 18. Mobius has moved for a preliminary injunction that enjoins Sun Nuclear from:
"(1) promoting, marketing, advertising, selling, and offering for sale [Sun Nuclear]'s ImagePro software product or software modules therein; and
(2) promoting, marketing, selling, and offering for sale any software products or modules derived through [Sun Nuclear]'s misappropriation of Mobius's trade secrets, regardless of whether these software products or modules are stand-alone products or incorporated into other products."
Doc. 7 at 20. It seeks preliminary injunctive relief from Sun Nuclear's alleged "breach of the Distribution and Transition Agreements, trade secret misappropriation, trade dress infringement and tortious interference with existing and prospective contract[s]." Doc. 7 at 19. For purposes of the instant motion, Mobius focused its request on its claim for trade secret misappropriation. Doc. 7 at 21.
A. Choice of Law
The Court must first determine whether Delaware law or Texas law governs the issues in this case. In diversity cases, federal courts apply the choice of law rules of the forum state. See Klaxon v. Stentor Elec. Mfg. Co., 313 U.S. 487, 496 (1941); Int'l Interests, L.P. v. Hardy, 448 F.3d 303, 306 (5th Cir. 2006) (citing Mayo v. Hartford Life Ins. Co., 354 F.3d 400, 403 (5th Cir. 2004)). Section 15.2 of the Distribution Agreement provides, "This Agreement shall be governed by and construed in accordance with the laws of the State of Delaware, U.S.A." Doc. 7, Ex. 3 ¶ 15.2. The Transition Agreement retained section 15.2 from the Distribution Agreement. Doc. 7, Ex. 4 ¶ 5. Accordingly, either the law of Texas or Delaware will apply to Mobius's claims.
Texas courts analyze the enforceability of choice of law provisions under section 187 of the Restatement (Second) of Conflict of Laws (1971). See DeSantis v. Wackenhut, Corp., 793 S.W.2d 670, 677-78 (Tex. 1990). "Section 187(1) allows the parties to incorporate by reference the laws of a forum to determine issues that could have been resolved by explicit agreement, such as "rules relating to construction" of an agreement." Caton v. Leach Corp., 896 F.2d 939, 942 (5th Cir. 1990) (citing Restatement (Second) of Conflict of Laws § 187(1) comment c (Supp.1988)); see also DeSantis, 793 S.W.2d at 677 ("When parties to a contract reside or expect to perform their respective obligations in multiple jurisdictions, they may be uncertain as to what jurisdiction's law will govern construction and enforcement of the contract. To avoid this uncertainty, they may express in their agreement their own choice that the law of a specified jurisdiction apply to their agreement. Judicial respect for their choice advances the policy of protecting their expectations.").
Section 187 states: Law of the State Chosen by the Parties
(1) The law of the state chosen by the parties to govern their contractual rights and duties will be applied if the particular issue is one which the parties could have resolved by an explicit provision in their agreement directed to that issue.
(2) The law of the state chosen by the parties to govern their contractual rights and duties will be applied, even if the particular issue is one which the parties could not have resolved by an explicit provision in their agreement directed to that issue, unless either
(a) the chosen state has no substantial relationship to the parties or the transaction and there is no other reasonable basis for the parties' choice, or(3) In the absence of a contrary indication of intention, the reference is to the local law of the state of the chosen law.
(b) application of the law of the chosen state would be contrary to a fundamental policy of a state which has a materially greater interest than the chosen state in the determination of the particular issue and which, under the rule of § 188, would be the state of the applicable law in the absence of an effective choice of law by the parties.
Because the parties chose to have the Agreements construed in accordance with Delaware law, the Court must give effect to that choice. See Fagan Holdings, Inc. v. Thinkware, Inc., 750 F. Supp. 2d 820, 825 (S.D. Tex. 2010) (holding that a claim for breach of contract should be governed by the law specified in the choice of law clause). However, in the instant case, the choice of law clause is narrowly written and does not govern the entire relationship of the parties. See Caton, 896 F.2d at 943 (finding that a similar, narrowly worded choice of law provision which denoted only that a particular state's law would be applied to "construe" the contract did not govern claims between the parties that arose outside the contract). Because Mobius has decided to focus its request for preliminary injunction on its claim for trade secret misappropriation, the Court must inquire further to determine what law should apply to govern this claim.
Under section 187(2) of the Restatement (Second), the parties' contractual choice of law will apply to any other claims as well unless 1) Delaware has no substantial relationship to the parties or the transaction and there is no reasonable basis for the parties' choice, or 2) another state has a materially greater interest than Delaware in determining the issue and that state's law would apply under section 188, in the absence of the parities' contrary choice. Restatement (Second) of Conflict of Laws § 188 (1971). From the record it appears that Delaware is an entirely arbitrary choice and that neither the parties nor the transaction bear any relationship whatsoever to Delaware. Sun Nuclear is organized under the laws of Florida, headquartered in Melbourne, Florida, and is engaged in business activities in Texas, among other jurisdictions. Doc. 1, Ex. 2 ¶ 4. Mobius is organized under the laws of Texas and headquartered in Bellaire, Texas. Doc. 1, Ex. 2 ¶ 3. In addition, DoseLab was developed in Texas. Pl.'s Reply at 4 (Doc. 22). Although institutions in Delaware may be using DoseLab, the Distribution Agreement made Sun Nuclear the exclusive worldwide distributor. As such, DoseLab presumably is used in hospitals and clinics worldwide and there does not appear any reason why Delaware would have a more significant relationship than any other jurisdiction wherein DoseLab is utilized in a hospital or clinic. The Court finds that Texas has a far more significant relationship to the parties and the agreement at issue and therefore holds that Texas law will apply to Mobius's trade secret claim.
Section 188 states: Law Governing in Absence of Effective Choice by the Parties
(1) The rights and duties of the parties with respect to an issue in contract are determined by the local law of the state which, with respect to that issue, has the most significant relationship to the transaction and the parties under the principles stated in § 6.
(2) In the absence of an effective choice of law by the parties (see § 187), the contacts to be taken into account in applying the principles of § 6 to determine the law applicable to an issue include:
(a) the place of contracting,These contacts are to be evaluated according to their relative importance with respect to the particular issue.
(b) the place of negotiation of the contract,
(c) the place of performance,
(d) the location of the subject matter of the contract, and
(e) the domicile, residence, nationality, place of incorporation and place of business of the parties.
(3) If the place of negotiating the contract and the place of performance are in the same state, the local law of this state will usually be applied, except as otherwise provided in §§ 189- 199 and 203. Restatement (Second) of Conflict of Laws § 188 (1971).
B. Legal Standard
To obtain a preliminary injunction, Mobius must show: (1) a substantial likelihood of success on the merits; (2) a substantial threat that it will suffer irreparable harm if the injunction is not granted; (3) that the threatened injury outweighs any damage that the injunction might cause the defendant; and (4) that the injunction will not disserve the public interest. Tex. Med. Providers Performing Abortion Servs. v. Lakey, 667 F.3d 570, 574 (5th Cir. 2012). Each of these factors presents a mixed question of fact and law. Opulent Life Church v. City of Holly Springs, Miss., 697 F.3d 279 (5th Cir. 2012). "A preliminary injunction is an 'extraordinary remedy' which should only be granted if the party seeking the injunction has 'clearly carried the burden of persuasion' on all four requirements." Nicols v. Alcatel USA, Inc., 532 F.3d 364, 372 (5th Cir. 2008). Under Texas law, "[t]he improper use of trade secrets provides a proper basis for an injunction." S.W. Research Inst. v. Keraplast Tech., Ltd., 103 S.W. 478, 482 (Tex, App.—San Antonio 2003, no pet.); see also Tex. Civ. Prac. Rem. Code § 134A.003 ("Actual or threatened misappropriation may be enjoined.").
C. Discussion
1. A Substantial Likelihood of Success on the Merits
"To state a claim for trade secret misappropriation under Texas law, a plaintiff must (1) establish that a trade secret existed; (2) demonstrate that the trade secret was acquired by the defendant through a breach of a confidential relationship or discovered by improper means; and (3) show that the defendant used the trade secret without authorization from the plaintiff." Gen. Univ. Sys. Inc. v. Hal, Inc., 379 F.3d 131, 149-50 (5th Cir. 2004). At the preliminary injunction stage, the trial court does not determine whether or not a trade secret actually exists, but only whether "the applicant has established that the information is entitled to trade secret protection until a trial on the merits." Fox v. Tropical Warehouses, Inc., 121 S.W.3d 853, 858 (Tex. App.—Fort Worth 2004, no pet.) (citing Ctr. For Econ. Justice v. Am. Ins. Ass'n, 39 S.W.3d 337, 343 (Tex. App.—Austin 2001, no pet.).
"A trade secret is any formula, pattern, device, or compilation of information used in one's business, and which gives one an opportunity to obtain an advantage over competitors who do not know or use it." Wellogix, Inc. v. Accenture, LLP, 823 F. Supp. 2d 555, 562 (citing Hyde Corp. v. Huffines, 158 Tex. 566, 314 S.W.2d 763, 776 (1958)). "To determine whether a trade secret exists, a court weighs six fact-intensive factors: (1) the extent to which the information is known outside of the business; (2) the extent to which it is known by employees and others involved in the business; (3) the extent of the measures taken to guard the secrecy of the information; (4) the value of the information to the business and to its competitors; (5) the amount of effort or money expended in developing the information; (6) the ease or difficulty with which the information could be properly acquired or duplicated by others." Id. (citing In re Bass, 113 S.W.3d 735, 739 (Tex. 2003)). The party claiming a trade secret need not satisfy all six factors "because trade secrets do not fit neatly into each factor every time." Gen. Univ. Sys., 379 F.3d 131, 150. The court must evaluate all of the factors in context. Id.
The parties vigorously dispute whether the alleged "trade secrets" at issue are contained in the user interface of DoseLab or in DoseLab's underlying algorithms and formulas. Mobius has stated repeatedly that it does not claim any trade secret protection in the user interface of DoseLab, but that DoseLab's "algorithms, visualization techniques, example files, configurations and parameters for its kV/MV imaging functionality, MLC log functionality, MLC strip test functionality, cone beam CT functionality, flatness and symmetry functionality, and starshot imaging" qualify as trade secrets under the Bass factors. See Pl.'s Reply at 8 (Doc. 22); Doc. 7 at 22.
The Court finds that for purposes of analyzing the trade secret issue at the preliminary injunction phase, Mobius has met its burden to show that at least some of the screenshots are representations of DoseLab's underlying proprietary trade secrets and not merely images of user interface. See Wellogix, Inc., 823 F. Supp. 2d at 562-63 (confirming that jury was entitled to credit expert's testimony that a software's user interface design "contained several pages of information that were trade secrets"). Even if some of Mobius's claimed trade secrets proved to be un-protectable "user interface" after a determination on the merits, that will not defeat Mobius's motion for a preliminary injunction. If any of the algorithms or formulas contained in ImagePro were misappropriated from DoseLab's trade secrets, Mobius is entitled to a preliminary injunction. "Where the defendant will be unable to prevent knowledge of plaintiff's trade secrets from infiltrating his work . . . Texas has recognized the need for injunctive relief . . . The fact that a single trade secret may be disclosed is enough." Picker Int'l, Inc. v. Blanton, 756 F. Supp. 971, 982 (N.D. Tex. 1990) (internal citations and quotation marks omitted). Therefore, the Court will consider under the Bass factors whether Mobius has met its burden to show that DoseLab's algorithms and formulas are entitled to trade secret protection pending a determination on the merits.
The Court finds that Mobius's algorithms and formulas are not well known outside its business. Sun Nuclear contends that the features and functionalities of DoseLab that ImagePro shares are either standard in the industry of TG-142 QA or are publically accessible through the open source version of DoseLab and through various journal articles. The Court disagrees. Although TG-142 QA is standard within the field of radiation oncology, DoseLab has offered sufficient evidence at this stage that it uses algorithms and formulas that are unique in the industry and which are not specifically provided in the TG-142 Report. See Wellogix, Inc., 823 F. Supp. 2d at 562-63 (confirming jury's verdict under Texas law when "the jury was presented with documentary evidence indicating that, though the functions of Wellogix's software were known to the industry, other software companies did not have identical functions in their software").
Furthermore, the algorithms and formulas in DoseLab are not well known by employees and others involved in Mobius's business. Dr. Childress testified that only five percent of the source code from the open source version of DoseLab was used in the first commercial version, and only three percent of the original source code is in the current version. Moreover, the open source version did not incorporate kV/MV imaging or cone beam CT imaging—two of the features which Mobius claims have been misappropriated. Dr. Childress also testified that those algorithms and formulas which he believes have been misappropriated have never appeared in any publication on TG-142 imaging.
The record shows that Mobius has taken considerable care to guard the secrecy of its confidential information. Dr. Childress testified that only five Mobius employees have access to DoseLab's source code, and those employees who do have access are required to sign a confidentiality agreement. Doc 7 at 23. Sun Nuclear was also bound by a confidentiality agreement under the terms of the Distribution Agreement, and access to Mobius's confidential information at Sun Nuclear was restricted to employees on a "need to know" basis. See Doc. 7, Ex. 3, ¶¶ 9.1-9.2. Also, with regard to DoseLab's users, Mobius requires both licensees and potential licensees to limit use and protect Mobius's confidential information via its End User License Agreement. Doc. 7 at 23. Based on these factors, the Court finds that DoseLab's trade secrets have not lost their secrecy through statements made to Sun Nuclear employees in the context of the distribution relationship or through demonstration or licensee use of DoseLab. See Metallurgical Indus. Inc. v. Fourtek, Inc., 790 F.2d 1195, 1200 (5th Cir. 1986) (finding no surrender of secrecy where disclosures were not public announcements and secrets divulged only to businesses with whom plaintiff dealt with expectation of profit).
The value of the trade secrets at issue in this case is not in reasonably in dispute. The algorithms and formulas for programming software that calibrates medical linear accelerators are extremely valuable to both Mobius and its competitors, including Sun Nuclear. Only three other companies produce the TG-142 solution software which is essential to maintaining QA for medical linear accelerators across the county and around the world. The trade secrets at issue are especially valuable to Mobius, as it is a young company which derives the vast majority of its income from sales of DoseLab. Doc. 7 at 23. The record reflects that Dr. Childress and his team at Mobius expended a considerable amount of time and resources to develop DoseLab Pro and improve upon its predecessor. The record also shows that Mobius's algorithms and formulas could not be easily acquired or duplicated by others. Mobius takes great precautions to guard its proprietary secrets. In addition, the high degree of technical knowledge of both medical physics and computer programming, which would be necessary to even understand DoseLab's algorithms and formulas, limits the ability of others to acquire or duplicate the information.
Examining the trade secrets at issue under the Bass factors clearly reveals that Mobius's proprietary information is entitled to trade secret protection until a determination on the merits. Mobius has also offered sufficient evidence that DoseLab's trade secrets were acquired through breach of a confidential relationship and were used without Mobius's authorization.
The Distribution Agreement itself provides clear prima facie evidence that the parties were in a confidential relationship which imposed a duty upon Sun Nuclear not to use Mobius's trade secrets. See IAC, Ltd. v. Bell Helicopter Textron, Inc., 160 S.W.3d 191, 199 (Tex. App.— Fort Worth 2007, pet. denied); see also Tex. Civ. Prac. Rem. Code § 134A.002(3) (defining "misappropriation"). By virtue of Sun Nuclear's distributorship, it had the opportunity to learn firsthand how DoseLab functioned. Dr. Childress testified that he shared details of DoseLab's algorithms and formulas with Dr. Daniels and other employees of Sun Nuclear to be used towards their shared goal of selling DoseLab licenses. Additionally, Sun Nuclear employees may have acquired information about DoseLab by downloading from Mobius's website 750 times during the development of ImagePro. Mobius argues, and the Court agrees, that Sun Nuclear's explanation that it was repeatedly downloading DoseLab for regular sales and marketing purposes is not credible because 1) Sun Nuclear was preparing to launch its competing TG-142 product within months of these downloads; and 2) Sun Nuclear's sales and customer support employees typically access Mobius's webserver from the customers' hospitals or clinics, not from the Sun Nuclear headquarters. Dr. Daniels testified that Sun Nuclear used its copy of DoseLab to verify that ImagePro was achieving accurate results, which was not an authorized use of DoseLab under the Distribution Agreement.
The Distribution Agreement prohibited Sun Nuclear from using or disclosing Mobius's confidential information, including its trade secrets. See Doc. 7, Ex. 3, ¶¶ 9.1-9.2. The replication of numerous functions and features of DoseLab in Sun Nuclear's ImagePro software provides sufficient evidence at this stage that Sun Nuclear "used" Mobius's trade secrets without its authorization. See Wellogix, 823 F. Supp. 2d at 566 ("'Use' of a trade secret means commercial use, by which a person seeks to profit from the use of the secret."). Based on these facts, the Court concludes that Mobius has met its burden to show a substantial likelihood of success on the merits of its trade secret misappropriation claim.
2. A Substantial Threat That Mobius Will Suffer Irreparable Harm
The Court finds that Mobius has met its burden to show that it will suffer irreparable harm if Sun Nuclear is not enjoined. Dr. Childress testified that during the course the parties' relationship, he shared confidential and proprietary information regarding DoseLab's formulas and algorithms with Dr. Daniels and other Sun Nuclear employees. Dr. Childress also testified that Sun Nuclear has shirked its remaining responsibilities as an international distributor of DoseLab under the Transition Agreement, and used its position in relation to existing DoseLab licensees to exploit consumer confusion to market ImagePro. "When a defendant possesses trade secrets and is in a position to use them, harm to the trade secret owner may be presumed." IAC, Ltd., 160 S.W. at 200 (internal citations omitted). Additionally, "considerable authority under Texas law indicat[es] that injuries to goodwill and competitive position are irreparable where trade secrets have been misappropriated..." Heil v. Trailer Intern. Co. v. Kula, 2013 WL 5630969, at * 4 (5th Cir. Oct. 16, 2013).
3. The Threatened Injury to Mobius Outweighs Any Damage the Injunction Might Cause Sun Nuclear
Mobius has met its burden to show that the balance of hardships weighs in its favor. Mobius was founded only a few years ago and sells a total of three products, including DoseLab. Doc. 7, Ex. 1 ¶¶. Its total revenues for 2012 were approximately $3 million, nearly all of which were derived from sales of DoseLab. Doc. 7, Ex. 1 ¶¶ 11, 42). For 2013, sales of DoseLab have comprised 79% of Mobius income to date. Doc 7, Ex. 1 ¶ 42. DoseLab is obviously critical to Mobius's existence. Sun Nuclear, by contrast, has revenues of approximately $60 million per year and sells over thirty products. See Doc. 14, Ex. 2. Sun Nuclear argues that it would be devastated by an injunction that prevents it from selling ImagePro, as sales of ImagePro are critical for it to recoup its investment of the human and financial resources expended in its development. Doc. 14 at 34. Sun Nuclear further argues that its reputation as a leader in the radiation oncology field would be harmed if it were unable to offer a TG-142 software solution to its customers. Doc. 14 at 34. Dr. Daniels testified that sales of its other products will suffer if it is unable to provide its own TG-142 QA software in a bundle of products because customers want to utilize one vendor for everything. Nevertheless, ImagePro has only been on the market for a few months, and Sun Nuclear has thrived nearly thirty years without offering its own TG-142 QA software solution. The Court finds that the balance of hardships clearly weigh in favor of Mobius.
4. The Injunction Will Not Disserve the Public Interest
There is no reason that the requested injunction would disserve the public interest. The injunction will not preclude Sun Nuclear from supporting its existing ImagePro customers who may have come to rely on that product, and there is no evidence that hospitals and clinics will be unable to rely on the currently existing and available QA software technology during the pendency of this litigation.
The Court finds that Mobius has met its burden of persuasion on all four requirements necessary to obtain the requested preliminary injunction.
III. Conclusion
For the foregoing reasons, it is hereby
ORDERED that Plaintiff Mobius Medical Systems, LP's request for preliminary injunction is GRANTED. It is further
ORDERED that Sun Nuclear Corporation is hereby enjoined, pending a final determination on the merits, from:
(1) promoting, marketing, advertising, selling, and offering for sale Sun Nuclear Corporation's ImagePro software products or software modules therein; and
(2) promoting, marketing, selling, and offering for sale any software products or modules derived through Sun Nuclear Corporation's misappropriation of Mobius Medical Systems, LP's trade secrets, regardless of whether these software products or modules are stand-alone products or incorporated into other products. It is further
ORDERED that the preliminary injunction issued contemporaneously herewith shall be effective upon the posting of a bond by Mobius Medical Systems, LP in the amount of twenty-five thousand dollars ($25,000 U.S.). FED. R. CIV. P. 65(c); Phillips v. Charles Schreiner Bank, 894 F.2d 127, 131 (5th Cir. 1990). The Court finds that this amount is appropriate to protect Sun Nuclear Corporation in the event that the injunction is later determined to be in error.
SIGNED at Houston, Texas, this 10th day of December, 2013.
_____________
MELINDA HARMON
UNITED STATES DISTRICT JUDGE