Receive a weekly summary and discussion of the top papers of the week by leading researchers in the field.

General General

ArXiv Preprint

Boolean Matrix Factorization (BMF) aims to find an approximation of a given binary matrix as the Boolean product of two low-rank binary matrices. Binary data is ubiquitous in many fields, and representing data by binary matrices is common in medicine, natural language processing, bioinformatics, computer graphics, among many others. Unfortunately, BMF is computationally hard and heuristic algorithms are used to compute Boolean factorizations. Very recently, the theoretical breakthrough was obtained independently by two research groups. Ban et al. (SODA 2019) and Fomin et al. (Trans. Algorithms 2020) show that BMF admits an efficient polynomial-time approximation scheme (EPTAS). However, despite the theoretical importance, the high double-exponential dependence of the running times from the rank makes these algorithms unimplementable in practice. The primary research question motivating our work is whether the theoretical advances on BMF could lead to practical algorithms. The main conceptional contribution of our work is the following. While EPTAS for BMF is a purely theoretical advance, the general approach behind these algorithms could serve as the basis in designing better heuristics. We also use this strategy to develop new algorithms for related $\mathbb{F}_p$-Matrix Factorization. Here, given a matrix $A$ over a finite field GF($p$) where $p$ is a prime, and an integer $r$, our objective is to find a matrix $B$ over the same field with GF($p$)-rank at most $r$ minimizing some norm of $A-B$. Our empirical research on synthetic and real-world data demonstrates the advantage of the new algorithms over previous works on BMF and $\mathbb{F}_p$-Matrix Factorization.

Fedor Fomin, Fahad Panolan, Anurag Patil, Adil Tanveer

2022-07-25

Surgery Surgery

Machine learning improves the accuracy of graft weight prediction in living donor liver transplantation.

In Liver transplantation : official publication of the American Association for the Study of Liver Diseases and the International Liver Transplantation Society

BACKGROUND : Precise graft weight estimation is essential for planning living donor liver transplantation to select grafts of adequate size for the recipient. This study aimed to investigate whether a machine learning model can improve the accuracy of graft weight estimation.

METHODS : Data from 872 consecutive living donors of a left lateral sector, left lobe, right lobe to adults or children for living-related liver transplantation were collected from January 2011 to December 2019. Supervised machine learning models were trained (80% of observations) to predict graft weight using the following information: donor's age, sex, height, weight, and BMI, graft type (left, right, or left lateral lobe), CT estimated graft and total liver volumes. Models' performance was measured in a random independent set (20% of observations) and in an external validation cohort using the mean absolute error and the mean absolute percentage error, and compared with methods currently available for graft weight estimation.

RESULTS : The best-performing machine learning model showed a mean absolute error value of 50 ± 62 grams in predicting graft weight, with a mean error of 10.3%. These errors were significantly lower than those observed with alternative methods. Additionally, 62% of predictions had errors <10%, while errors>15% were observed in only 18.4% of the cases, compared to the 34.6% of the predictions obtained with the best alternative method (p<0.001). The machine learning model is made available as a web application (http://graftweight.shinyapps.io/prediction).

CONCLUSIONS : Machine learning can improve the precision of graft weight estimation compared to currently available methods by reducing the frequency of significant errors. The coupling of anthropometric variables to the preoperatively estimated graft volume seems necessary to improve the accuracy of graft weight estimation.

Giglio Mariano Cesare, Zanfardino Mario, Franzese Monica, Zakaria Hazem, Alobthani Salah, Zidan Ahmed, Ayoub Islam Ismail, Shoreem Hany Abdelmeguid, Lee Boram, Han Ho-Seong, Della Penna Andrea, Nadalin Silvio, Troisi Roberto Ivan, Broering Dieter Clement

2022-Sep-27

Radiology Radiology

Ageing and degeneration analysis using ageing-related dynamic attention on lateral cephalometric radiographs.

In NPJ digital medicine

With the increase of the ageing in the world's population, the ageing and degeneration studies of physiological characteristics in human skin, bones, and muscles become important topics. Research on the ageing of bones, especially the skull, are paid much attention in recent years. In this study, a novel deep learning method representing the ageing-related dynamic attention (ARDA) is proposed. The proposed method can quantitatively display the ageing salience of the bones and their change patterns with age on lateral cephalometric radiographs images (LCR) images containing the craniofacial and cervical spine. An age estimation-based deep learning model based on 14142 LCR images from 4 to 40 years old individuals is trained to extract ageing-related features, and based on these features the ageing salience maps are generated by the Grad-CAM method. All ageing salience maps with the same age are merged as an ARDA map corresponding to that age. Ageing salience maps show that ARDA is mainly concentrated in three regions in LCR images: the teeth, craniofacial, and cervical spine regions. Furthermore, the dynamic distribution of ARDA at different ages and instances in LCR images is quantitatively analyzed. The experimental results on 3014 cases show that ARDA can accurately reflect the development and degeneration patterns in LCR images.

Zhang Zhiyong, Liu Ningtao, Guo Zhang, Jiao Licheng, Fenster Aaron, Jin Wenfan, Zhang Yuxiang, Chen Jie, Yan Chunxia, Gou Shuiping

2022-Sep-27

General General

Tracking the Body, Wing, and Leg Kinematics of Moving Mosquitoes.

In Cold Spring Harbor protocols

In this protocol, we discuss general techniques for tracking the three-dimensional (3D) locations of the mosquito body, wings, legs, or other features of interest using videos. Tracking data must be acquired to produce detailed kinematics of moving mosquitoes. The software of focus for this protocol, DLTdv, was chosen for its widespread use and excellent support and because it is open-source. In addition, DLTdv allows both manual and automatic tracking. The automatic tracking can be done using a classic machine vision or machine-learning algorithm. The software supports both single-camera analysis and multicamera systems and can take advantage of sophisticated calibration algorithms, both for intrinsic lens distortion correction and for 3D DLT-based reconstruction. For this protocol, we assume all kinematic data is acquired post hoc through video analysis.

Hedrick Tyson L, Dickerson Andrew K, Muijres Florian T, Pieters Remco

2022-Sep-28

Ophthalmology Ophthalmology

Automatic interpretation and clinical evaluation for fundus fluorescein angiography images of diabetic retinopathy patients by deep learning.

In The British journal of ophthalmology

BACKGROUND/AIMS : Fundus fluorescein angiography (FFA) is an important technique to evaluate diabetic retinopathy (DR) and other retinal diseases. The interpretation of FFA images is complex and time-consuming, and the ability of diagnosis is uneven among different ophthalmologists. The aim of the study is to develop a clinically usable multilevel classification deep learning model for FFA images, including prediagnosis assessment and lesion classification.

METHODS : A total of 15 599 FFA images of 1558 eyes from 845 patients diagnosed with DR were collected and annotated. Three convolutional neural network (CNN) models were trained to generate the label of image quality, location, laterality of eye, phase and five lesions. Performance of the models was evaluated by accuracy, F-1 score, the area under the curve and human-machine comparison. The images with false positive and false negative results were analysed in detail.

RESULTS : Compared with LeNet-5 and VGG16, ResNet18 got the best result, achieving an accuracy of 80.79%-93.34% for prediagnosis assessment and an accuracy of 63.67%-88.88% for lesion detection. The human-machine comparison showed that the CNN had similar accuracy with junior ophthalmologists. The false positive and false negative analysis indicated a direction of improvement.

CONCLUSION : This is the first study to do automated standardised labelling on FFA images. Our model is able to be applied in clinical practice, and will make great contributions to the development of intelligent diagnosis of FFA images.

Gao Zhiyuan, Pan Xiangji, Shao Ji, Jiang Xiaoyu, Su Zhaoan, Jin Kai, Ye Juan

2022-Sep-28

Imaging, Retina, Telemedicine