SpikeDeep-Classifier: A deep-learning based fully automatic offline spike sorting algorithm

The most widely used spike-sorting algorithms are semiautomatic in practice, requiring manual tuning of the automatic solution to achieve good performance. In this work, we propose a new fully automatic spike-sorting algorithm that can capture multiple clusters of different sizes and densities. In addition, we introduce an improved feature selection method, by using a variable number of wavelet coefficients, based on the degree of non-Gaussianity of their distributions. We evaluated the performance of the proposed algorithm with real and simulated data. With real data from single-channel recordings, in ~95% of the cases the new algorithm replicated, in an unsupervised way, the solutions obtained by expert sorters, who manually optimized the solution of a previous semiautomatic algorithm. This was done while maintaining a low number of false positives. With simulated data from single-channel and tetrode recordings, the new algorithm was able to correctly detect many more neurons compared with previous implementations and also compared with recently introduced algorithms, while significantly reducing the number of false positives. In addition, the proposed algorithm showed good performance when tested with real tetrode recordings. NEW & NOTEWORTHY We propose a new fully automatic spike-sorting algorithm, including several steps that allow the selection of multiple clusters of different sizes and densities. Moreover, it defines the dimensionality of the feature space in an unsupervised way. We evaluated the performance of the algorithm with real and simulated data, from both single-channel and tetrode recordings. The proposed algorithm was able to outperform manual sorting from experts and other recent unsupervised algorithms.

Download Full-text

A Fully Automatic Deep Learning System for L3 Slice Selection and Body Composition Assessment on Abdominal Computed Tomography. (Preprint)

10.2196/preprints.28399 ◽

2021 ◽

Author(s):

Jiyeon Ha ◽

Taeyong Park ◽

Hong-Kyu Kim ◽

Youngbin Shin ◽

Yousun Ko ◽

...

Keyword(s):

Computed Tomography ◽

Deep Learning ◽

External Validation ◽

Abdominal Muscle ◽

Anatomic Variations ◽

Success Rates ◽

Technical Success ◽

Segmentation Accuracy ◽

Distance Difference ◽

Fully Automatic

BACKGROUND As sarcopenia research has been gaining emphasis, the need for quantification of abdominal muscle on computed tomography (CT) is increasing. Thus, a fully automated system to select L3 slice and segment muscle in an end-to-end manner is demanding. OBJECTIVE We aimed to develop a deep learning model (DLM) to select the L3 slice with consideration of anatomic variations and to segment cross-sectional areas (CSAs) of abdominal muscle and fat. METHODS Our DLM, named L3SEG-net, was composed of a YOLOv3-based algorithm for selecting the L3 slice and a fully convolutional network (FCN)-based algorithm for segmentation. The YOLOv3-based algorithm was developed via supervised learning using a training dataset (n=922), and the FCN-based algorithm was transferred from prior work. Our L3SEG-net was validated with internal (n=496) and external validation (n=586) datasets. L3 slice selection accuracy was evaluated by the distance difference between ground truths and DLM-derived results. Technical success for L3 slice selection was defined when the distance difference was <10 mm. Overall segmentation accuracy was evaluated by CSA error. The influence of anatomic variations on DLM performance was evaluated. RESULTS In the internal and external validation datasets, the accuracy of automatic L3 slice selection was high, with mean distance differences of 3.7±8.4 mm and 4.1±8.3 mm, respectively, and with technical success rates of 93.1% and 92.3%, respectively. However, in the subgroup analysis of anatomic variations, the L3 slice selection accuracy decreased, with distance differences of 12.4±15.4 mm and 12.1±14.6 mm, respectively, and with technical success rates of 67.2% and 67.9%, respectively. The overall segmentation accuracy of abdominal muscle areas was excellent regardless of anatomic variation, with the CSA errors of 1.38–3.10 cm2. CONCLUSIONS A fully automatic system was developed for the selection of an exact axial CT slice at the L3 vertebral level and the segmentation of abdominal muscle areas.

Download Full-text

A fully automatic deep learning system for L3 slice selection and body composition assessment on abdominal computed tomography.

10.21203/rs.3.rs-598394/v1 ◽

2021 ◽

Author(s):

Jiyeon Ha ◽

Taeyong Park ◽

Hong-Kyu Kim ◽

Youngbin Shin ◽

Yousun Ko ◽

...

Keyword(s):

Computed Tomography ◽

Deep Learning ◽

External Validation ◽

Abdominal Muscle ◽

Anatomic Variations ◽

Success Rates ◽

Technical Success ◽

Segmentation Accuracy ◽

Distance Difference ◽

Fully Automatic

Abstract Background and aims: As sarcopenia research has been gaining emphasis, the need for quantification of abdominal muscle on computed tomography (CT) is increasing. Thus, a fully automated system to select L3 slice and segment muscle in an end-to-end manner is demanded. We aimed to develop a deep learning model (DLM) to select the L3 slice with consideration of anatomic variations and to segment cross-sectional areas (CSAs) of abdominal muscle and fat. Methods: Our DLM, named L3SEG-net, was composed of a YOLOv3-based algorithm for selecting the L3 slice and a fully convolutional network (FCN)-based algorithm for segmentation. The YOLOv3-based algorithm was developed via supervised learning using a training dataset (n=922), and the FCN-based algorithm was transferred from prior work. Our L3SEG-net was validated with internal (n=496) and external validation (n=586) datasets. L3 slice selection accuracy was evaluated by the distance difference between ground truths and DLM-derived results. Technical success for L3 slice selection was defined when the distance difference was <10 mm. Overall segmentation accuracy was evaluated by CSA error. The influence of anatomic variations on DLM performance was evaluated.Results: In the internal and external validation datasets, the accuracy of automatic L3 slice selection was high, with mean distance differences of 3.7±8.4 mm and 4.1±8.3 mm, respectively, and with technical success rates of 93.1% and 92.3%, respectively. However, in the subgroup analysis of anatomic variations, the L3 slice selection accuracy decreased, with distance differences of 12.4±15.4 mm and 12.1±14.6 mm, respectively, and with technical success rates of 67.2% and 67.9%, respectively. The overall segmentation accuracy of abdominal muscle areas was excellent regardless of anatomic variation, with the CSA errors of 1.38–3.10 cm2.Conclusions: A fully automatic system was developed for the selection of an exact axial CT slice at the L3 vertebral level and the segmentation of abdominal muscle areas.

Download Full-text

Fully automatic segmentation of LV from Echocardiography images and calculation of Ejection Fraction using Deep Learning

International Journal of Biomedical Engineering and Technology ◽

10.1504/ijbet.2020.10036183 ◽

2020 ◽

Vol 1 (1) ◽

pp. 1

Author(s):

Deepa Madathil ◽

Pallavi Kulkarni

Keyword(s):

Deep Learning ◽

Ejection Fraction ◽

Automatic Segmentation ◽

Fully Automatic

Download Full-text

Automatic spike sorting for high-density microelectrode arrays

Journal of Neurophysiology ◽

10.1152/jn.00803.2017 ◽

2018 ◽

Vol 120 (6) ◽

pp. 3155-3171 ◽

Cited By ~ 13

Author(s):

Roland Diggelmann ◽

Michele Fiscella ◽

Andreas Hierlemann ◽

Felix Franke

Keyword(s):

Principal Component Analysis ◽

Clustering Algorithm ◽

State Of The Art ◽

Microelectrode Arrays ◽

Principal Component ◽

Component Analysis ◽

High Density ◽

Sorting Algorithm ◽

Spike Sorting ◽

Recording Electrodes

High-density microelectrode arrays can be used to record extracellular action potentials from hundreds to thousands of neurons simultaneously. Efficient spike sorters must be developed to cope with such large data volumes. Most existing spike sorting methods for single electrodes or small multielectrodes, however, suffer from the “curse of dimensionality” and cannot be directly applied to recordings with hundreds of electrodes. This holds particularly true for the standard reference spike sorting algorithm, principal component analysis-based feature extraction, followed by k-means or expectation maximization clustering, against which most spike sorters are evaluated. We present a spike sorting algorithm that circumvents the dimensionality problem by sorting local groups of electrodes independently with classical spike sorting approaches. It is scalable to any number of recording electrodes and well suited for parallel computing. The combination of data prewhitening before the principal component analysis-based extraction and a parameter-free clustering algorithm obviated the need for parameter adjustments. We evaluated its performance using surrogate data in which we systematically varied spike amplitudes and spike rates and that were generated by inserting template spikes into the voltage traces of real recordings. In a direct comparison, our algorithm could compete with existing state-of-the-art spike sorters in terms of sensitivity and precision, while parameter adjustment or manual cluster curation was not required. NEW & NOTEWORTHY We present an automatic spike sorting algorithm that combines three strategies to scale classical spike sorting techniques for high-density microelectrode arrays: 1) splitting the recording electrodes into small groups and sorting them independently; 2) clustering a subset of spikes and classifying the rest to limit computation time; and 3) prewhitening the spike waveforms to enable the use of parameter-free clustering. Finally, we combined these strategies into an automatic spike sorter that is competitive with state-of-the-art spike sorters.

Download Full-text