Department of Medicine, Division of Cardiology Bakar Computational Health Sciences Institute UCSF - UC Berkeley Joint Program in Computational Precision Health Department of Radiology, Center for Intelligent Imaging University of California, San Francisco Corresponding Author Keywords: medical imaging, video translation, deep learning, image synthesis, ultrasound Word Count: 4129 Abstract Deep learning (DL) has the potential to revolutionize image acquisition and interpretation across medicine, h owever, attention to data imbalance and missin gness is required . U ltrasound data presents a particular challenge because in addition to different views and structures, it includes several sub - modalities -- such as greyscale and color flow doppler (CFD) -- that are often imbalanced in clinical studies . Image translation can help balance datasets but is challenging for ultrasound sub - modalities to date . Here, we present a generative method for ultrasound CFD - greyscale video translation, t rained on 5 4, 975 videos and tested on 8, 3 68 . The method developed leveraged pixel - wise, adversarial, and perceptual loses and utilized two networks: one for reconstructing anatomic structures and one for denoising to achieve realistic ultrasound imaging . A verage pairwise SSIM between synthetic videos and ground truth was 0.9 1 0.0 4 . Synthetic videos performed indistinguishably from real ones in DL classification and segmentation tasks and when evaluated by b linded clinical experts: F1 score was 0.9 for real and 0.89 for synthetic videos; Dice score between real and synthetic segmentation was 0.97. Overall c linician accuracy in distinguishing real vs synthetic videos was 54 6% (42 - 61%), indicating reali stic synthetic videos . Although trained only on heart videos, the model worked well on ultrasound spanning several clinical domains (av erage SSIM 0.91 0.0 5), demonstrating foundational abilit ies .

Tabular data dominates data science but poses challenges for generative models, especially when the data is limited or sensitive. We present a novel approach to generating synthetic tabular data based on the principle of maximum entropy -- MaxEnt -- called GEM-T, for ``generative entropy maximization for tables.'' GEM-T directly captures nth-order interactions -- pairwise, third-order, etc. -- among columns of training data. In extensive testing, GEM-T matches or exceeds deep neural network approaches previously regarded as state-of-the-art in 23 of 34 publicly available datasets representing diverse subject domains (68\%). Notably, GEM-T involves orders-of-magnitude fewer trainable parameters, demonstrating that much of the information in real-world data resides in low-dimensional, potentially human-interpretable correlations, provided that the input data is appropriately transformed first. Furthermore, MaxEnt better handles heterogeneous data types (continuous vs. discrete vs. categorical), lack of local structure, and other features of tabular data. GEM-T represents a promising direction for light-weight high-performance generative models for structured data.

Generative models hold great potential, but only if one can trust the evaluation of the data they generate. We show that many commonly used quality scores for comparing two-dimensional distributions of synthetic vs. ground-truth data give better results than they should, a phenomenon we call the "grade inflation problem." We show that the correlation score, Jaccard score, earth-mover's score, and Kullback-Leibler (relative-entropy) score all suffer grade inflation. We propose that any score that values all datapoints equally, as these do, will also exhibit grade inflation; we refer to such scores as "equipoint" scores. We introduce the concept of "equidensity" scores, and present the Eden score, to our knowledge the first example of such a score. We found that Eden avoids grade inflation and agrees better with human perception of goodness-of-fit than the equipoint scores above. We propose that any reasonable equidensity score will avoid grade inflation. We identify a connection between equidensity scores and R\'enyi entropy of negative order. We conclude that equidensity scores are likely to outperform equipoint scores for generative models, and for comparing low-dimensional distributions more generally.

The University of California--Irvine (UCI) Machine Learning (ML) Repository (UCIMLR) is consistently cited as one of the most popular dataset repositories, hosting hundreds of high-impact datasets. However, a significant portion, including 28.4% of the top 250, cannot be imported via the $\textit{ucimlrepo}$ package that is provided and recommended by the UCIMLR website. Instead, they are hosted as .zip files, containing nonstandard formats that are difficult to import without additional ad hoc processing. To address this issue, here we present $\textit{lucie}$ -- $\underline{l}oad$ $\underline{U}niversity$ $\underline{C}alifornia$ $\underline{I}rvine$ $\underline{e}xamples$ -- a utility that automatically determines the data format and imports many of these previously non-importable datasets, while preserving as much of a tabular data structure as possible. $\textit{lucie}$ was designed using the top 100 most popular datasets and benchmarked on the next 130, where it resulted in a success rate of 95.4% vs. 73.1% for $\textit{ucimlrepo}$. $\textit{lucie}$ is available as a Python package on PyPI with 98% code coverage.

In deep learning, achieving high performance on image classification tasks requires diverse training sets. However, the current best practice$\unicode{x2013}$maximizing dataset size and class balance$\unicode{x2013}$does not guarantee dataset diversity. We hypothesized that, for a given model architecture, model performance can be improved by maximizing diversity more directly. To test this hypothesis, we introduce a comprehensive framework of diversity measures from ecology that generalizes familiar quantities like Shannon entropy by accounting for similarities among images. (Size and class balance emerge as special cases.) Analyzing thousands of subsets from seven medical datasets showed that the best correlates of performance were not size or class balance but $A$$\unicode{x2013}$"big alpha"$\unicode{x2013}$a set of generalized entropy measures interpreted as the effective number of image-class pairs in the dataset, after accounting for image similarities. One of these, $A_0$, explained 67% of the variance in balanced accuracy, vs. 54% for class balance and just 39% for size. The best pair of measures was size-plus-$A_1$ (79%), which outperformed size-plus-class-balance (74%). Subsets with the largest $A_0$ performed up to 16% better than those with the largest size (median improvement, 8%). We propose maximizing $A$ as a way to improve deep learning performance in medical imaging.

Artificial intelligence is being used increasingly in medicine to make services more efficient and improve patient care. Now, local researchers plan to use the technology to perform "virtual biopsies" of brain tumors. Steven Hibbert considered himself a healthy guy. Then suddenly his life changed on a dime. "I was just kind of sitting there, reading the paper, and went into a seizure," said Hibbert. The 55-year-old Cape Cod resident was rushed to the hospital and got the news no one wants to hear.

Artificial intelligence is being used increasingly in medicine to make services more efficient and improve patient care. Now, researchers in Boston plan to use the technology to perform "virtual biopsies" of brain tumors. The goal is to help patients like Steven Hibbert, who considered himself a healthy guy until one day life changed on a dime. "I was just kind of sitting there, reading the paper, and went into a seizure," Hibbert told CBS Boston's Dr. Mallika Marshall. The 55-year-old Cape Cod resident was rushed to the hospital and got the news no one wants to hear. He had a brain tumor.

Artificial intelligence is already set to affect countless areas of your life, from your job to your health care. New research reveals it could soon be used to analyze your heart. AI could soon be used to analyze your heart. A study published Wednesday found that advanced machine learning is faster, more accurate and more efficient than board-certified echocardiographers at classifying heart anatomy shown on an ultrasound scan. The study was conducted by researchers from the University of California, San Francisco, the University of California, Berkeley, and Beth Israel Deaconess Medical Center.

Rima Arnaout, an assistant professor, and cardiologist at UC San Francisco, is working on her research in computational medicine; she published a new study in the journal Digital Medicine. In the study, Arnaout and her colleagues used deep learning, specifically something called a convolutional neural network (CNN), to train an Artificial Intelligence system that can classify echocardiograms. The system is created to analyze heart scan, that is just a simple task. The system is to outperform the human cardiologists but not to replace them. It was a limited task, she notes, just the first step in what a cardiologist does when evaluating an echocardiogram (the image produced by bouncing sound waves off the heart).

A scientist in the US has developed a "fast" and "accurate" artificial intelligence system that can classify echocardiogram results using deep-learning algorithm. Rima Arnaout, from the University of California San Francisco, said the AI system is capable of analysing heart scans faster and more accurately than human cardiologists. However, speaking to IEEE Spectrum, she warned that the technology won't replace doctors, but could help double-check scans, speed-up evaluation and, perhaps, highlight factors the doctors may have missed. Currently, the tool is limited because it can only evaluate echocardiograms - ultrasounds of the heart. "The best technique is still inside the head of the trained echocardiographer," she said.