Digital avatars of patients lungs

Medi­cine isn’t just about drug treat­ments. Dif­fe­rences bet­ween patients are not limi­ted to gene­tics. At MΞDISIM, resear­chers are ana­ly­sing the bio­me­cha­nics of disea­sed tis­sue to create digi­tal models for indi­vi­dual patients as a way bet­ter gui­ding the­ra­peu­tic choices.

“The lung changes shape when we breathe, and doubles in volume”, Cécile Patte, a resear­cher and reci­pient of the 2020 L’O­réal-UNES­CO for Women in Science Award who recent­ly defen­ded her doc­to­ral the­sis at MΞDISIM, points out. “The way it changes shape depends on its mecha­ni­cal pro­per­ties, such as elas­ti­ci­ty.” She deve­lo­ped a digi­tal lung model in order to help doc­tors at Avi­cenne Hos­pi­tal in France to bet­ter unders­tand the indi­vi­dual cha­rac­te­ris­tics of patients with idio­pa­thic pul­mo­na­ry fibro­sis – a chro­nic lung disease, which is one of the long-term effects of Covid-19.

In this condi­tion, scar tis­sue forms in the lungs, ren­de­ring them stiff, blo­cking the pas­sage of oxy­gen to the blood and lea­ding to fatal res­pi­ra­to­ry fai­lure within just two to five years. Signi­fi­cant­ly, the shape, poro­si­ty and mecha­ni­cal pro­per­ties of the lung are not the same for all patients. “In order to ans­wer gene­ral ques­tions, you only need a model of an ave­rage lung, using ave­rage mea­su­re­ments for the mecha­ni­cal pro­per­ties. But for a per­so­na­li­sed approach, you need to build an ava­tar of the patient’s organ,” she explains. This means acqui­ring data.

Real life data

“We want to work with exis­ting data, although it makes things more dif­fi­cult. We don’t want to sub­ject patients to extra inva­sive pro­ce­dures in order to create the ava­tar,” Patte says. The models are the­re­fore based on data col­lec­ted during the nor­mal course of the patient’s treat­ment. Exams vary depen­ding on the kind of patho­lo­gy and the organ in ques­tion. For the heart, there are blood pres­sure and elec­tro­phy­sio­lo­gy mea­su­re­ments. For the lung, it’s main­ly ima­ging tech­niques, X‑rays and scans. “If we could mea­sure pul­mo­na­ry pleu­ral pres­sure, for example, the model would be more accu­rate, but the test is too inva­sive.”

The patient’s spe­ci­fic infor­ma­tion is fed into a digi­tal ava­tar, enabling more detai­led diag­no­sis, which includes quan­ti­fying mecha­ni­cal para­me­ters, eva­lua­ting poten­tial treat­ments and pre­dic­ting prog­no­sis. “This infor­ma­tion should help doc­tors bet­ter diag­nose the disease and guide the patient towards drug the­ra­py or transplantation.”

Cur­rent­ly, per­so­na­li­sed simu­la­tion is only at the research, or ‘proof of concept’, stage. Much work still needs to be done before it can be applied cli­ni­cal­ly and used as a medi­cal device to help make the­ra­peu­tic deci­sions. The regu­la­to­ry envi­ron­ment is strict and sub­ject to EU requi­re­ments, such as CE marking.

So far, Patte’s lung simu­la­tion soft­ware has only been applied to four patients. “Using the com­pu­ting capa­ci­ty of the lab’s ser­ver clus­ter, our soft­ware needs up to three full days to pro­cess one patient’s data,” she admits. Pro­ces­sing time will have to be shor­te­ned for the sys­tem to work in cli­ni­cal practice.

A digi­tal twin

The lung is not the only organ the team is model­ling : a lot of work is being done on the heart with mul­tiple cli­ni­cal appli­ca­tions, and in direct col­la­bo­ra­tion with various hos­pi­tals. The aim is to represent the heart and its various cha­rac­te­ris­tics, inclu­ding size, contrac­ti­li­ty and elec­tro­phy­sio­lo­gy. Per­so­na­li­sed car­diac models could enable doc­tors to test out a treat­ment vir­tual­ly, par­ti­cu­lar­ly in the case of heart fai­lure, and pre­dict whe­ther or not the patient would respond.

MΞDISIM has also desi­gned a tool, Anaes­tAs­sist, to moni­tor the car­dio­vas­cu­lar sys­tem in real time while a patient is anes­the­ti­sed. This soft­ware models the patient’s car­dio­vas­cu­lar phy­sio­lo­gy in order to pre­dict the effects of the anaes­the­tic and assist the doc­tor during the ope­ra­tion. Other labo­ra­to­ries throu­ghout the world are wor­king on model­ling blood flow, bones, kid­neys, etc. Patte believes that “all organs can be stu­died with this approach.”

It may one day be pos­sible to bring all these models toge­ther in order to simu­late sys­te­mic effects and create full digi­tal twins of patients to help pin­point indi­vi­dua­li­sed treat­ments. This idea has even gai­ned ground in indus­trial circles, with invest­ment from com­pa­nies such as Das­sault Sys­tèmes, who crea­ted a plat­form cal­led 3DEXPERIENCE that can help medi­cal device manu­fac­tu­rers opti­mize pro­duct deve­lop­ment through simulation.