arXiv subjects

Laurent Milot

Publications and source records attributed to Laurent Milot.

4 recordsLinked to original sources

arXiv2025

Interstitial dual-mode ultrasound with a 3-mm MR-compatible catheter for image-guided HIFU and directional in-vitro tissue ablations

Current interstitial techniques of tumor ablation face challenges that ultrasound technologies could meet. The ablation radius and directionality of the ultrasound beam could improve the efficiency and precision. Here, a 9gauge MR-compatible dual-mode ultrasound catheter prototype was experimentally evaluated for Ultrasound Imageguided High Intensity Focused Ultrasound (USgHIFU) conformal ablations. The prototype consisted of 64 piezocomposite linear array elements and was driven by an open research programmable dual-mode ultrasound platform. After verifying the US-image guidance capabilities of the prototype, the HIFU output performances (dynamic focusing and HIFU intensities) were quantitatively characterized, together with the associated 3D HIFU-induced thermal heating in tissue phantoms (using MR thermometry). Finally, the ability to produce robustly HIFU-induced thermal ablations in in-vitro liver was studied experimentally and compared to numerical modeling. Investigations of several HIFU dynamic focusing allowed overcoming the challenges of miniaturizing the device: mono-focal focusing maximized deep energy deposition, while multi-focal strategies eliminated grating lobes. The linear-array design of the prototype made it possible to produce interstitial ultrasound images of tissue and tumor mimics in situ. Multi-focal pressure fields were generated without grating lobes and transducer surface intensities reached up to Isapa =14 W$\bullet$cm -2 . Seventeen elementary thermal ablations were performed in vitro. Rotation of the catheter proved the directionality of ablation, sparing non-targeted tissue. This experimental proof of concept demonstrates the feasibility of treating volumes comparable to those of primary solid tumors with a miniaturized USgHIFU catheter whose dimensions are close to those of tools traditionally used in interventional radiology, while offering new functionalities.

physics.class-ph↗

arXiv2021

AMINN: Autoencoder-based Multiple Instance Neural Network Improves Outcome Prediction of Multifocal Liver Metastases

Colorectal cancer is one of the most common and lethal cancers and colorectal cancer liver metastases (CRLM) is the major cause of death in patients with colorectal cancer. Multifocality occurs frequently in CRLM, but is relatively unexplored in CRLM outcome prediction. Most existing clinical and imaging biomarkers do not take the imaging features of all multifocal lesions into account. In this paper, we present an end-to-end autoencoder-based multiple instance neural network (AMINN) for the prediction of survival outcomes in multifocal CRLM patients using radiomic features extracted from contrast-enhanced MRIs. Specifically, we jointly train an autoencoder to reconstruct input features and a multiple instance network to make predictions by aggregating information from all tumour lesions of a patient. Also, we incorporate a two-step normalization technique to improve the training of deep neural networks, built on the observation that the distributions of radiomic features are almost always severely skewed. Experimental results empirically validated our hypothesis that incorporating imaging features of all lesions improves outcome prediction for multifocal cancer. The proposed AMINN framework achieved an area under the ROC curve (AUC) of 0.70, which is 11.4% higher than the best baseline method. A risk score based on the outputs of AMINN achieved superior prediction in our multifocal CRLM cohort. The effectiveness of incorporating all lesions and applying two-step normalization is demonstrated by a series of ablation studies. A Keras implementation of AMINN is released.

cs.CV↗

arXiv2021

Comparing Deep Learning strategies for paired but unregistered multimodal segmentation of the liver in T1 and T2-weighted MRI

We address the problem of multimodal liver segmentation in paired but unregistered T1 and T2-weighted MR images. We compare several strategies described in the literature, with or without multi-task training, with or without pre-registration. We also compare different loss functions (cross-entropy, Dice loss, and three adversarial losses). All methods achieved comparable performances with the exception of a multi-task setting that performs both segmentations at once, which performed poorly.

eess.IV↗

arXiv2019

Unsupervised Clustering of Quantitative Imaging Phenotypes using Autoencoder and Gaussian Mixture Model

Quantitative medical image computing (radiomics) has been widely applied to build prediction models from medical images. However, overfitting is a significant issue in conventional radiomics, where a large number of radiomic features are directly used to train and test models that predict genotypes or clinical outcomes. In order to tackle this problem, we propose an unsupervised learning pipeline composed of an autoencoder for representation learning of radiomic features and a Gaussian mixture model based on minimum message length criterion for clustering. By incorporating probabilistic modeling, disease heterogeneity has been taken into account. The performance of the proposed pipeline was evaluated on an institutional MRI cohort of 108 patients with colorectal cancer liver metastases. Our approach is capable of automatically selecting the optimal number of clusters and assigns patients into clusters (imaging subtypes) with significantly different survival rates. Our method outperforms other unsupervised clustering methods that have been used for radiomics analysis and has comparable performance to a state-of-the-art imaging biomarker.

eess.IV↗