报告题目：Prediction of fertilization outcome by electrophysiological and mechanical quantification of the uterine activity enhanced by machine learning

报 告 人：Massimo Mischi

报告时间：2023年10月14日 14:00-16:00

报告地点：计算机A401

报告摘要：

  About 20% of women in reproductive age have difficulty to get pregnant, with in-vitrofertilization (IVF) representing the last-resort treatment. Besides the high cost, IVF success rates remain below 30%. In IVF, after hormonal stimulation, the   produced oocytes are picked up, fertilized in vitro, and the formed embryo(s) are transferred back into the uterine cavity. There is increasing evidence that uterine receptivity, supported by   favorable uterine peristalsis (UP), plays an important role in successful fertilization (embryo implantation). 

Therefore, over the past 10 years we have extensively investigated the uterine activity. We first developed dedicated algorithms for the analysis of multichannel surface electrohysterography (EHG),registering the electromyographic activity of the uterus. We were able to detect changes during the different phases of the menstrual cycle and to extract EHG features capable to predict IVF success prior to the embryo transfer into the uterine cavity. We then developed dedicated solutions for uterine strain analysis by ultrasound speckle tracking, adaptively adjusted to the uterine anatomy. We complemented the estimation of strain frequency and amplitude features with novel features based on UP velocity, direction,and coordination, extracted by dedicated spatiotemporal analysis through k-space representations of the measured, local strains. In both EHG and ultrasound recordings, singular value decomposition techniques were employed for denoising as well as to produce additional features providing a global representation of the uterine activity. A machine-learning (ML) framework was then developed to optimally combine the available complementary informationwith the aim of boosting the prediction accuracy of the IVF outcome.

Patient data have been acquired in several European medical centers from 123 patients. During this presentation, the results obtained for the prediction of IVF success by employment of different individual features, as well as by their combination through a ML prediction model, will be discussed. Dedicated prediction models have been developed and tested for individual modalities (EHG and ultrasound) as well as for their multimodal combination. Besides the prediction of IVF outcome, the developed tools for uterine quantitative analysis have the potential to aid in the diagnosis of common uterine problems such as adenomyosis and myomas.

报告人简介：

Massimo Mischi received an MSc in Electronic Engineering at La Sapienza University in Rome (Italy, 1999). He then received a PhD degree at the Eindhoven University of Technology (Netherlands, 2004), Electrical Engineering department, working on cardiovascular quantification by contrast-enhanced ultrasound imaging. Massimo Mischi is now professor of biomedical signal analysis in the same department, where he leads the Signal Processing Systems Unit and the Biomedical Diagnostics lab. This lab counts over 80 researchers working on model-based quantitative analysis of biosignals, ranging from electrophysiology to diagnostic imaging. Artificial intelligence is further investigated to boost the interpretation accuracy of biosignals by integration of physics-driven and data-driven approaches. Especially his research on ultrasound imaging of angiogenesis was pluri-awarded with several personal grants (VIDI, ERC Starting Grant, ERC Proof of Concept). He has also contributed to the valorization of his scientific output with the foundation of two start-up companies in the fields of neuromuscular rehabilitation (HiPerMotion) and prostate cancer diagnostics (Angiogenesis Analytics). Overall, he contributed to over 180 peer-reviewed journal papers, 13 book chapters, one book, 15 patents, and over 70 invited talks at international conferences such as AIUM, EUROSON (keynote), IEEE IUS, and EAU. He currently serves as an associate editor for the IEEE T-UFFC, IEEE RBME, CMPB, Sensors, and IRBM. He is also chairman of the IEEE EMBS Benelux Chapter, a board member of the Urological Imaging Section of the European Association of Urology, a member of the Safety Committee of the World Federation of Ultrasound in Medicine and Biology, and secretary of the Dutch Society of Medical Ultrasound.