Shaping ping the Future ure of Biomedic medical al Device e - PowerPoint PPT Presentation

Shaping ping the Future ure of Biomedic medical al Device e Design ign and Diagnos gnostic tics Kristia ian Debus us

Overvie Ov view Biomed omedica cal l Device ice Design: gn: Bencht nchtop op and Clini nical cal Trial Sup uppor ort – Simulation and the regulatory system • ASME V&V 40, MDDT, MDIC, Avicenna – Case Studies: the path forward • VEXTEC Corp and CD-adapco: drug delivery in the cardiovascular system • Simulia Living Heart Project • System Integration: Lumped Parameter Model Optimization with Optimate+ Biomed omedica cal l Diagno nost stic ics s – Business Concept Options: Lean Startup Concepts, The “Uber Model”, Reimbursement – inFluids LLC

Biomedical Device Design: The Possibilities • Regulatory and Industry converging on solutions through consortia work Directions and • “Diagnostics Simulations” put new Trends business concepts to the test • A shared vision of Computational Clinical Trials is evolving • New to Simulation: Understand the Challenges benefits, needs and pain?  Evangelism • Already using Simulation: Transition from Benchtop to Clinical Trials • Work with Consortia, Academics, Marketing Goal: Accelerate • Collaborate with Diagnostic Tool Makers Industry Acceptance • Demonstrate Clinical Trial Project Success / Go Beyond Simulation

Rel elevance ce of Sim imula lati tion on and d Mode delin ing g to Supp pport t Clin inic ical Tria ials • Device ice developme elopment nt and modeli odeling ng un under er consideration nsideration of – Variation of patient specific data – Product variations – Variation of patient/environment conditions: Rest, Exercise, Blood Pressure, Drug Effects etc. • CD CD-adapco adapco memb mber er of AS ASME E V&V-40 40, , MDIC IC

ASME E V&V 40, MDDT, , MDIC, C, Avicenn nna

MDIC, C, ASME E & FDA

AVICE CENN NNA Avicenna: A Strategy for in silico Clinical Trials Tasked by the European Commission (EC) to produce a Roadmap for the introduction of in silico clinical trials, the Avicenna project began in October 2013 and runs until September 2015. The project will develop and promote this Roadmap, and work to overcome the legal, financial, organizational and technical barriers that could slow the adoption of computer simulation in this domain. http://avicenna-isct.org/

The FDA’s MDDT - Medical Device Development Tools • The FDA’s Medical Device Development Tools (MDDT) program is a way for the FDA to qualify tools that medical device sponsors can use in the development and evaluation of medical devices. • Qualification means that the FDA has evaluated the tool and concurs with available supporting evidence that the tool produces scientifically-plausible measurements and works as intended within the specified context of use …

Case e Studies: udies: VEXTE TEC C Corp Objective: Simulate the motion of a large number of interacting particles through the human vasculature – Example: loaded drug-eluting beads being delivered to a target site as treatment Key features – Tracking the particles in a numerically efficient manner – Modeling contact forces and interactions – Considering stochastic effects associated with random distributions of particle size and number – Goal: Determine the impact of above factors and variability of inlet flow conditions on the effectiveness of the therapy  the coverage fraction by the particles with the target vasculature.

Case e Studies: udies: VEXTE TEC C Corp STAR-CCM+ used to model flow of particles within the vessels along with the interactions VEXTEC's VLM Uncertainty Management platform handles random statistical nature of the problem to generate the uncertainty scenarios

Dassault sault/Simu Simulia: lia: The Living ing Hear art t Project oject

Workflow: Biomedical Device Design using Patient Specific Data  Simulating Systems Analysis & Geometry CFD / FEA / FSI MRI/CT/PET Visualization Generation Simulation Scan • Segmentation of DICOM Sets (Image Slices) • CAD or STAR-CCM+ Optimization Integration of Device Bench Top Comparison (Stent Graft) and Patient Data FEA solution provided by Simulia DICOM segmentation provided by Materialise STAR-CCM+ Workflow Capabilities: STL Cleanup – Java (Windkessel) – Simulation Assistant - Optimization

Medical al De Device e De Design ign and nd Syst stem em Integr grat ation: ion: Lumped Parameter Model Optimization with Optimate+ The e gold d st standa ndard d for haemod emodynam ynamic ic si simulation ulations is s the e us use of lumped umped parame meter er mode dels s to desc scribe ribed d the e downs wnstr tream eam impedance mpedance of the e vascula ulatu ture. re. – The most commonly used lumped parameter model is the three element Windkessel which is constructed from two resistors in series and a capacitor in parallel. A key A k y challenge lenge to us using ng a W Windk ndkess ssel el model odel (or any ot other her lum umped ped model) odel) is the e appropriat priate e choice oice of parame meter value ues. s. – Common approaches are: • Simplified analytical calculations – often good to get realistic initial values but generally not suitable for “matched” or production cases. • Manual trial and error – time consuming and generally only gets you so far. • Optimisation procedure – automated process which should obtain the “best” fit

Optim imat ate+: +: Pa Paramet eter er defini nitio tion When hen run unnin ning g an optimisat imisation ion study udy we have to define ine parame meter ers s which ich can be modi odified ied by the e algori rith thm m and an object jectiv ive(s e(s) ) which ch is s to be minimis imised ed or maximi imise sed. d. Parame Par meter ers – These are simply the Windkessel parameters, Z, R and C. – S ince these parameters are defined in a java macro, which can’t be “seen” by Optimate+, we have to modify our approach to allow Optimate+ access. • This can be done by creating a field function for each of the Windkessel parameters and modifying the java to read the value of these field functions rather than setting them explicitly in the java macro

Optim imat ate+: +: Pa Paramet eter er defini nitio tion Parame Par meter er defini initi tion on – Z is allowed to vary between 9e6 to 1.3e7 with a resolution of 51 • Baseline value of 1.1e7 – R is allowed to vary between 1.15e8 to 1.75e8 with a resolution of 51 • Baseline value of 1.45e8 – C is allowed to vary between 1.1e-8 to 2.1e-8 with a resolution of 201 • Baseline value of 1.45e-8 The e baseli eline ne value ues s had d already ady been en manually ly tun uned ed to match ch the clinical cal wavef eform orm in previ vious ous wor ork. – Comparison of baseline values • Green = Clinical Pressure • Red = Windkessel Pressure

Optim imat ate+: +: Results ts Variati tion on in Windk ndkess ssel el press ssures ures

Optim imat ate+: +: Results ts Comparison of the clinical, baseline, and “best” fit pressure waveforms

Biomed omedical ical Diagnost nostic ics: s: Busin siness ess Concept ept Options ions Lean n Star artups tups Rei eimbur mbursem ement ent Strat rategy: egy: Hear artf tflow The “ Uberiza erizati tion ” of Healt althcare hcare ( St Stuar uart t Karten en - founder nder of Karten en Design) n) – Just as Uber changed transportation …., healthcare will be infiltrated by startups wanting to change the healthcare model from hospital-centric to patient-centric. Medical device companies and other healthcare providers that don’t realize that a major shift is taking place will become the equivalent of today’s taxi industry….. – We are also seeing a transformation of the man-machine relationship — we are starting to wear computers, and soon, we will be implanting them into our bodies to connect with our communication systems, cars, and homes. As artificial intelligence improves, it will help us interact with increasingly smart environments. In healthcare, highly evolved sensors and powerful algorithms will give us proactive, personalized care. By 2035, the majority of our treatments will occur at home. Our home will be watching us and helping us track our health ….. – We are e alrea eady dy seeing eing more empo mpower ered ed patien ients. ts. Peop ople le wa want infor ormation. mation. They wa want t to make their ir own diagn agnosis osis. . They wa want to resea search their ir doc octor ors. . They wa want their ir own healt alth h data….. – It all starts now. Healthcare must shift its focus toward the patient. Successful medical products will put the patient’s needs first and foremost. …. Impa mpact ct on Simulat ation ion Market? t?

Biom omed edica ical l Diagnostic nostics: s: inFluidS luidS LLC “Saving lives, one simulation at a time ” ™ At InFluidS, we bring clarity and insight to pulmonary physician’s most complex diagnostic challenges. shahab.taherian@csulb.edu