Using machine learning to understand neuromorphological change and image-based biomarker identification in Cavalier King Charles Spaniels with Chiari-like malformation-associated pain and syringomyelia

Artificial intelligence in veterinary diagnostic imaging: Perspectives and limitations

The field of veterinary diagnostic imaging is undergoing significant transformation with the integration of artificial intelligence (AI) tools. This manuscript provides an overview of the current state and future prospects of AI in veterinary diagnostic imaging. The manuscript delves into various applications of AI across different imaging modalities, such as radiology, ultrasound, computed tomography, and magnetic resonance imaging. Examples of AI applications in each modality are provided, ranging from orthopaedics to internal medicine, cardiology, and more. Notable studies are discussed, demonstrating AI's potential for improved accuracy in detecting and classifying various abnormalities. The ethical considerations of using AI in veterinary diagnostics are also explored, highlighting the need for transparent AI development, accurate training data, awareness of the limitations of AI models, and the importance of maintaining human expertise in the decision-making process. The manuscript underscores the significance of AI as a decision support tool rather than a replacement for human judgement. In conclusion, this comprehensive manuscript offers an assessment of the current landscape and future potential of AI in veterinary diagnostic imaging. It provides insights into the benefits and challenges of integrating AI into clinical practice while emphasizing the critical role of ethics and human expertise in ensuring the wellbeing of veterinary patients.

Machine learning algorithms predict canine structural epilepsy with high accuracy

Introduction

Clinical reasoning in veterinary medicine is often based on clinicians' personal experience in combination with information derived from publications describing cohorts of patients. Studies on the use of scientific methods for patient individual decision making are largely lacking. This applies to the prediction of the individual underlying pathology in seizuring dogs as well. The aim of this study was to apply machine learning to the prediction of the risk of structural epilepsy in dogs with seizures.

Materials and methods

Dogs with a history of seizures were retrospectively as well as prospectively included. Data about clinical history, neurological examination, diagnostic tests performed as well as the final diagnosis were collected. For data analysis, the Bayesian Network and Random Forest algorithms were used. A total of 33 features for Random Forest and 17 for Bayesian Network were available for analysis. The following four feature selection methods were applied to select features for further analysis: Permutation Importance, Forward Selection, Random Selection and Expert Opinion. The two algorithms Bayesian Network and Random Forest were trained to predict structural epilepsy using the selected features.

Results

A total of 328 dogs of 119 different breeds were identified retrospectively between January 2017 and June 2021, of which 33.2% were diagnosed with structural epilepsy. An overall of 89,848 models were trained. The Bayesian Network in combination with the Random feature selection performed best. It was able to predict structural epilepsy with an accuracy of 0.969 (sensitivity: 0.857, specificity: 1.000) among all dogs with seizures using the following features: age at first seizure, cluster seizures, seizure in last 24 h, seizure in last 6 month, and seizure in last year.

Conclusion

Machine learning algorithms such as Bayesian Networks and Random Forests identify dogs with structural epilepsy with a high sensitivity and specificity. This information could provide some guidance to clinicians and pet owners in their clinical decision-making process.

Is foramen magnum decompression for acquired Chiari I malformation like putting a finger in the dyke? - A simplistic overview of artificial intelligence in assessing critical upstream and downstream etiologies

Background

Missed diagnosis of evolving or coexisting idiopathic (IIH) and spontaneous intracranial hypotension (SIH) is often the reason for persistent or worsening symptoms after foramen magnum decompression for Chiari malformation (CM) I. We explore the role of artificial intelligence (AI)/convolutional neural networks (CNN) in Chiari I malformation in a combinatorial role for the first time in literature, exploring both upstream and downstream magnetic resonance findings as initial screening profilers in CM-1. We have also put together a review of all existing subtypes of CM and discuss the role of upright (gravity-aided) magnetic resonance imaging (MRI) in evaluating equivocal tonsillar descent on a lying-down MRI. We have formulated a workflow algorithm MaChiP 1.0 (Manjila Chiari Protocol 1.0) using upstream and downstream profilers, that cause de novo or worsening Chiari I malformation, which we plan to implement using AI.

Materials and Methods

The PRISMA guidelines were used for "CM and machine learning and CNN" on PubMed database articles, and four articles specific to the topic were encountered. The radiologic criteria for IIH and SIH were applied from neurosurgical literature, and they were applied between primary and secondary (acquired) Chiari I malformations. An upstream etiology such as IIH or SIH and an isolated downstream etiology in the spine were characterized using the existing body of literature. We propose the utility of using four selected criteria for IIH and SIH each, over MRI T2 images of the brain and spine, predominantly sagittal sequences in upstream etiology in the brain and multiplanar MRI in spinal lesions.

Results

Using MaChiP 1.0 (patent/ copyright pending) concepts, we have proposed the upstream and downstream profilers implicated in progressive Chiari I malformation. The upstream profilers included findings of brain sagging, slope of the third ventricular floor, pontomesencephalic angle, mamillopontine distance, lateral ventricular angle, internal cerebral vein-vein of Galen angle, and displacement of iter, clivus length, tonsillar descent, etc., suggestive of SIH. The IIH features noted in upstream pathologies were posterior flattening of globe of the eye, partial empty sella, optic nerve sheath distortion, and optic nerve tortuosity in MRI. The downstream etiologies involved spinal cerebrospinal fluid (CSF) leak from dural tear, meningeal diverticula, CSF-venous fistulae, etc.

Conclusion

AI would help offer predictive analysis along the spectrum of upstream and downstream etiologies, ensuring safety and efficacy in treating secondary (acquired) Chiari I malformation, especially with coexisting IIH and SIH. The MaChiP 1.0 algorithm can help document worsening of a previously diagnosed CM-1 and find the exact etiology of a secondary CM-I. However, the role of posterior fossa morphometry and cine-flow MRI data for intracranial CSF flow dynamics, along with advanced spinal CSF studies using dynamic myelo-CT scanning in the formation of secondary CM-I is still being evaluated.

The effect of MRI-based screening and selection on the prevalence of syringomyelia in the Dutch and Danish Cavalier King Charles Spaniels

Introduction

Syringomyelia (SM) is a heritable disorder causing a fluid filled cavity (FFC) in the spinal cord with a reported overall prevalence of 39 to 46% in the Cavalier King Charles Spaniels (CKCS). Breeders started screening their CKCS with MRI in the Netherlands since 2004 and in Denmark since 2015. The goal of this study was to evaluate the effect of MRI-based selection in breeding on the prevalence of SM.

Method

MRI scans of 2,125 purebred CKCS were available. SM was defined as having a visible FFC in the spinal cord. The prevalence of SM per year of birth was calculated, and a logistic regression was used to evaluate the affected status of offspring from affected versus unaffected parents and age category of the parent and study the combined effect of parental status and age-category to evaluate the effect on the affected status of the offspring.

Results

The mean FFC in affected CKCS was 2.03 ± 1.47 mm and ranged from 0.5 to 9 mm (median of 1.5 mm). An age effect exists as older CKCS, which has a higher frequency of being affected compared with younger CKCS. There was no significant sex predilection for SM in this dataset. The mean prevalence of SM decreased slightly from 38% (2010-2014; 2.8 ± 1.3 years of age (mean ± sd); median 2.6 years) to 27% (2015-2019; 2.4 ± 1.2 years of age; median 2.1 years) in the screened population of CKCS (p = 4.3e-07). Breeding with two affected parents increased the odds ratio with 3.08 for producing affected offspring (95% CI 1.58-6.04) compared with breeding with unaffected parents.

Discussion

MRI-based screening and selection against SM led to a minimal decrease in the prevalence of SM in the Dutch and Danish CKCS population. Breeding with dogs with SM significantly increases the risk of affected offspring. As the disorder is progressive with age, and based on the results of this study, MRI-based screening for all CKCS is recommended at an age of 3 years or older, and to reduce SM more effectively, CKCS affected with SM should not be used for breeding.

Phenotypic characterization of Pomeranians with or without Chiari-like malformation and syringomyelia

Introduction

Chiari-like malformation (CM) and syringomyelia (SM) are frequently diagnosed conditions in small and toy dog breeds, such as the Cavalier King Charles Spaniel and Griffon Bruxellois. CM/SM is only rarely reported in Pomeranians in literature to date. The aims of this study are to 1/describe the phenotype of Pomeranians with or without CM/SM and 2/evaluate for differences and associations between CM/SM and owner-reported clinical signs (ORCS) or signalment factors.

Materials and methods

From February 2015 to June 2023, historical data and signalment (including country of origin, pedigree, sex and neuter status, age, and body weight) and ORCS of Pomeranians were recorded at multiple institutions. MRI studies of all dogs were evaluated for classification of CM/SM. Additionally, quantitative measurements were performed for SM.

Results

A total of 796 dogs from 22 different countries were included. Total prevalence of CM was 54.9% (437/796) and the prevalence of SM was 23.9% (190/796). The top 5 ORCS included 1/scratching with skin contact, rubbing head or ears, or both (57.6% of dogs with ORCS), 2/air licking (30.7% of dogs with ORCS), 3/spontaneous signs of pain (26.0% of dogs with ORCS), 4/persistent licking front and/or hind paws (22.6% of dogs with ORCS), 5/phantom scratching (22.6% of dogs with ORCS). Phantom scratching, vocalization, head shaking, spontaneous signs of pain, and air licking were associated with having SM. There were no statistically significant associations between quantitative syrinx measurements and ORCS. There were statistically significant associations between CM classification and 1/country of origin, 2/having a pedigree, and 3/age. There were statistically significant associations between SM classification and 1/age and 2/body weight.

Discussion

This is the first large study evaluating CM/SM in the Pomeranian dog breed. Veterinary clinicians can use these findings to increase the likelihood of correctly determining the presence or absence of CM/SM in Pomeranians. Breeders may consider using the information regarding signalment factors as well as ORCS associated with CM/SM classifications to select dogs for screening procedures. But an MRI-based diagnosis is needed to properly ascertain the exact CM/SM status of their breeding stock until a fool-proof characteristic or genetic marker is found.

Artificial Intelligence in Veterinary Imaging: An Overview

Artificial intelligence and machine learning have been increasingly used in the medical imaging field in the past few years. The evaluation of medical images is very subjective and complex, and therefore the application of artificial intelligence and deep learning methods to automatize the analysis process would be very beneficial. A lot of researchers have been applying these methods to image analysis diagnosis, developing software capable of assisting veterinary doctors or radiologists in their daily practice. This article details the main methodologies used to develop software applications on machine learning and how veterinarians with an interest in this field can benefit from such methodologies. The main goal of this study is to offer veterinary professionals a simple guide to enable them to understand the basics of artificial intelligence and machine learning and the concepts such as deep learning, convolutional neural networks, transfer learning, and the performance evaluation method. The language is adapted for medical technicians, and the work already published in this field is reviewed for application in the imaging diagnosis of different animal body systems: musculoskeletal, thoracic, nervous, and abdominal.

Artificial intelligence in veterinary diagnostic imaging: A literature review

Artificial intelligence in veterinary medicine is an emerging field. Machine learning, a subfield of artificial intelligence, allows computer programs to analyze large imaging datasets and learn to perform tasks relevant to veterinary diagnostic imaging. This review summarizes the small, yet growing body of artificial intelligence literature in veterinary imaging, provides necessary background to understand these papers, and provides author commentary on the state of the field. To date, less than 40 peer-reviewed publications have utilized machine learning to perform imaging-associated tasks across multiple anatomic regions in veterinary clinical and biomedical research. Major challenges in this field include collection and cleaning of sufficient image data, selection of high-quality ground truth labels, formation of relationships between veterinary and machine learning professionals, and closure of the gap between academic uses of artificial intelligence and currently available commercial products. Further development of artificial intelligence has the potential to help meet the growing need for radiological services through applications in workflow, quality control, and image interpretation for both general practitioners and radiologists.

Evaluating artificial intelligence algorithms for use in veterinary radiology

Artificial intelligence is increasingly being used for applications in veterinary radiology, including detection of abnormalities and automated measurements. Unlike human radiology, there is no formal regulation or validation of AI algorithms for veterinary medicine and both general practitioner and specialist veterinarians must rely on their own judgment when deciding whether or not to incorporate AI algorithms to aid their clinical decision-making. The benefits and challenges to developing clinically useful and diagnostically accurate AI algorithms are discussed. Considerations for the development of AI research projects are also addressed. A framework is suggested to help veterinarians, in both research and clinical practice contexts, assess AI algorithms for veterinary radiology.

The unmet potential of artificial intelligence in veterinary medicine

Veterinary medicine is a broad and growing discipline that includes topics such as companion animal health, population medicine and zoonotic diseases, and agriculture. In this article, we provide insight on how artificial intelligence works and how it is currently applied in veterinary medicine. We also discuss its potential in veterinary medicine. Given the rapid pace of research and commercial product developments in this area, the next several years will pose challenges to understanding, interpreting, and adopting this powerful and evolving technology. Artificial intelligence has the potential to enable veterinarians to perform tasks more efficiently while providing new insights for the management and treatment of disorders. It is our hope that this will translate to better quality of life for animals and those who care for them.

Using Deep Learning to Detect Spinal Cord Diseases on Thoracolumbar Magnetic Resonance Images of Dogs

Deep Learning based Convolutional Neural Networks (CNNs) are the state-of-the-art machine learning technique with medical image data. They have the ability to process large amounts of data and learn image features directly from the raw data. Based on their training, these networks are ultimately able to classify unknown data and make predictions. Magnetic resonance imaging (MRI) is the imaging modality of choice for many spinal cord disorders. Proper interpretation requires time and expertise from radiologists, so there is great interest in using artificial intelligence to more quickly interpret and diagnose medical imaging data. In this study, a CNN was trained and tested using thoracolumbar MR images from 500 dogs. T1- and T2-weighted MR images in sagittal and transverse planes were used. The network was trained with unremarkable images as well as with images showing the following spinal cord pathologies: intervertebral disc extrusion (IVDE), intervertebral disc protrusion (IVDP), fibrocartilaginous embolism (FCE)/acute non-compressive nucleus pulposus extrusion (ANNPE), syringomyelia and neoplasia. 2,693 MR images from 375 dogs were used for network training. The network was tested using 7,695 MR images from 125 dogs. The network performed best in detecting IVDPs on sagittal T1-weighted images, with a sensitivity of 100% and specificity of 95.1%. The network also performed very well in detecting IVDEs, especially on sagittal T2-weighted images, with a sensitivity of 90.8% and specificity of 98.98%. The network detected FCEs and ANNPEs with a sensitivity of 62.22% and a specificity of 97.90% on sagittal T2-weighted images and with a sensitivity of 91% and a specificity of 90% on transverse T2-weighted images. In detecting neoplasms and syringomyelia, the CNN did not perform well because of insufficient training data or because the network had problems differentiating different hyperintensities on T2-weighted images and thus made incorrect predictions. This study has shown that it is possible to train a CNN in terms of recognizing and differentiating various spinal cord pathologies on canine MR images. CNNs therefore have great potential to act as a “second eye” for imagers in the future, providing a faster focus on the altered image area and thus increasing workflow in radiology.

Association between filum terminale internum length and pain in Cavalier King Charles spaniels with and without syringomyelia

Background

Lumbar syringomyelia (SM), lumbosacral pain, and more caudal spinal cord termination are reported in Cavalier King Charles spaniels (CKCS). Data are lacking on the clinical relevance of alterations in their spinal cord terminal structures.

Objectives

To compare spinal cord termination level and filum terminale internum length (FTIL) with presence of lumbar SM and clinical signs in CKCS.

Animals

Forty‐eight CKCS.

Methods

In this prospective study, pain was quantified using owner and clinician assessments. Vertebral level of spinal cord and dural sac termination, presence of SM, and FTIL were determined from sagittal magnetic resonance imaging (MRI) sequences. Kappa and intraclass correlation (ICC) analyses determined interobserver reliability. The MRI findings were compared to owner and clinician‐reported pain quantification.

Results

Interobserver reliability was good for spinal cord and dural sac termination (kappa = 0.61 and 0.64, respectively) and excellent for FTIL (ICC: 92% agreement). The spinal cord terminated at 6th lumbar vertebra in 1, 7th lumbar vertebra in 31, and the sacrum in 15 dogs, and termination level was associated with lumbar SM (P = .002) but not clinical signs. Mean FTIL was 2.9 ± 1.08 mm; it was associated with owner‐reported pain (P = .033) and spinal palpation scores (P = .023). Painful CKCS without SM had shorter FTIL compared to normal CKCS and painful CKCS with SM (P = .02).

Conclusions

Painful CKCS without SM have decreased distance between the termination of the spinal cord and dural sac, suggesting a shorter FTIL. More caudal spinal cord termination is associated with development of lumbar SM.

Using machine learning to understand neuromorphological change and image-based biomarker identification in Cavalier King Charles Spaniels with Chiari-like malformation-associated pain and syringomyelia

Background

Chiari‐like malformation (CM) is a complex malformation of the skull and cranial cervical vertebrae that potentially results in pain and secondary syringomyelia (SM). Chiari‐like malformation‐associated pain (CM‐P) can be challenging to diagnose. We propose a machine learning approach to characterize morphological changes in dogs that may or may not be apparent to human observers. This data‐driven approach can remove potential bias (or blindness) that may be produced by a hypothesis‐driven expert observer approach.

Hypothesis/Objectives

To understand neuromorphological change and to identify image‐based biomarkers in dogs with CM‐P and symptomatic SM (SM‐S) using a novel machine learning approach, with the aim of increasing the understanding of these disorders.

Animals

Thirty‐two client‐owned Cavalier King Charles Spaniels (CKCSs; 11 controls, 10 CM‐P, 11 SM‐S).

Methods

Retrospective study using T2‐weighted midsagittal Digital Imaging and Communications in Medicine (DICOM) anonymized images, which then were mapped to images of an average clinically normal CKCS reference using Demons image registration. Key deformation features were automatically selected from the resulting deformation maps. A kernelized support vector machine was used for classifying characteristic localized changes in morphology.

Results

Candidate biomarkers were identified with receiver operating characteristic curves with area under the curve (AUC) of 0.78 (sensitivity 82%; specificity 69%) for the CM‐P biomarkers collectively and an AUC of 0.82 (sensitivity, 93%; specificity, 67%) for the SM‐S biomarkers, collectively.

Conclusions and clinical importance

Machine learning techniques can assist CM/SM diagnosis and facilitate understanding of abnormal morphology location with the potential to be applied to a variety of breeds and conformational diseases.