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.