A method for labelling lesions for machine learning and some new observations on osteochondrosis in computed tomographic scans of four pig joints.

BACKGROUND: Osteochondrosis is a major cause of leg weakness in pigs. Selection against osteochondrosis is currently based on manual scoring of computed tomographic (CT) scans for the presence of osteochondrosis manifesta lesions. It would be advantageous if osteochondrosis could be diagnosed automatically, through artificial intelligence methods using machine learning. The aim of this study was to describe a method for labelling articular osteochondrosis lesions in CT scans of four pig joints to guide development of future machine learning algorithms, and to report new observations made during the labelling process. The shoulder, elbow, stifle and hock joints were evaluated in CT scans of 201 pigs. RESULTS: Six thousand two hundred fifty osteochondrosis manifesta and cyst-like lesions were labelled in 201 pigs representing a total volume of 211,721.83 mm3. The per-joint prevalence of osteochondrosis ranged from 64.7% in the hock to 100% in the stifle joint. The lowest number of lesions was found in the hock joint at 208 lesions, and the highest number of lesions was found in the stifle joint at 4306 lesions. The mean volume per lesion ranged from 26.21 mm3 in the shoulder to 42.06 mm3 in the elbow joint. Pigs with the highest number of lesions had small lesions, whereas pigs with few lesions frequently had large lesions, that have the potential to become clinically significant. In the stifle joint, lesion number had a moderate negative correlation with mean lesion volume at r = - 0.54, p < 0.001. CONCLUSIONS: The described labelling method is an important step towards developing a machine learning algorithm that will enable automated diagnosis of osteochondrosis manifesta and cyst-like lesions. Both lesion number and volume should be considered during breeding selection. The apparent inverse relationship between lesion number and volume warrants further investigation.

A high-throughput study of visceral organs in CT-scanned pigs.

It has been debated whether intensive selection for growth and carcass yield in pig breeding programmes can affect the size of internal organs, and thereby reduce the animal's ability to handle stress and increase the risk of sudden deaths. To explore the respiratory and circulatory system in pigs, a deep learning based computational pipeline was built to extract the size of lungs and hearts from CT-scan images. This pipeline was applied on CT images from 11,000 boar selection candidates acquired during the last decade. Further, heart and lung volumes were analysed genetically and correlated with production traits. Both heart and lung volumes were heritable, with h2 estimated to 0.35 and 0.34, respectively, in Landrace, and 0.28 and 0.4 in Duroc. Both volumes were positively correlated with lean meat percentage, and lung volume was negatively genetically correlated with growth (rg = - 0.48 ± 0.07 for Landrace and rg = - 0.44 ± 0.07 for Duroc). The main findings suggest that the current pig breeding programs could, as an indirect response to selection, affect the size of hearts- and lungs. The presented methods can be used to monitor the development of internal organs in the future.

Osteochondrosis and other lesions in all intervertebral, articular process and rib joints from occiput to sacrum in pigs with poor back conformation, and relationship to juvenile kyphosis.

BACKGROUND: Computed tomography (CT) is used to evaluate body composition and limb osteochondrosis in selection of breeding boars. Pigs also develop heritably predisposed abnormal curvature of the spine including juvenile kyphosis. It has been suggested that osteochondrosis-like changes cause vertebral wedging and kyphosis, both of which are identifiable by CT. The aim of the current study was to examine the spine from occiput to sacrum to map changes and evaluate relationships, especially whether osteochondrosis caused juvenile kyphosis, in which case CT could be used in selection against it. Whole-body CT scans were collected retrospectively from 37 Landrace or Duroc boars with poor back conformation scores. Spine curvature and vertebral shape were evaluated, and all inter-vertebral, articular process and rib joints from the occiput to the sacrum were assessed for osteochondrosis and other lesions. RESULTS: Twenty-seven of the 37 (73%) pigs had normal spine curvature, whereas 10/37 (27%) pigs had abnormal curvature and all of them had wedge vertebrae. The 37 pigs had 875 focal lesions in articular process and rib joints, 98.5% of which represented stages of osteochondrosis. Five of the 37 pigs had focal lesions in other parts of vertebrae, mainly consisting of vertebral body osteochondrosis. The 10 pigs with abnormal curvature had 21 wedge vertebrae, comprising 10 vertebrae without focal lesions, six ventral wedge vertebrae with ventral osteochondrosis lesions and five dorsal wedge vertebrae with lesions in the neuro-central synchondrosis, articular process or rib joints. CONCLUSIONS: Computed tomography was suited for identification of wedge vertebrae, and kyphosis was due to ventral wedge vertebrae compatible with heritably predisposed vertebral body osteochondrosis. Articular process and rib joint osteochondrosis may represent incidental findings in wedge vertebrae. The role of the neuro-central synchondrosis in the pathogenesis of vertebral wedging warrants further investigation.

Computed tomographic development of physeal osteochondrosis in pigs.

BACKGROUND: Articular osteochondrosis follows a dynamic development pattern. Lesions arise, in incidence peaks compatible with failure of cartilage canal vessels during incorporation into bone, and can also resolve. Lesions that resolve before examination at a single time point will constitute false-negative diagnoses. The aim of the study was to identify physeal osteochondrosis lesions in pigs and monitor their development by computed tomography (CT), to determine if they follow a similar dynamic development pattern to articular osteochondrosis. RESULTS: Thirteen physes were evaluated bilaterally in up to eight biweekly CT scans from 18 male Landrace pigs age 70-180 days (total: 112 scans), generating 2912 scores. There were 1754 (60%) lesion-negative scores and 1158 (40%) lesion-positive scores. Positive scores comprised 138 lesions present at the start and 235 lesions that developed during the study, from 4 to 32 lesions per physis (median: 15 lesions). There were 1-2 peaks in the incidence curves for 12/13 examined physes, the exception being the proximal humerus. Positive scores also included 785 times that lesions persisted, from 1.3-4.8 examination intervals per lesion (median: 2.8 intervals). Negative scores included 190 times that lesions resolved, from 19 to 100% of lesions per physis (median: 65%). Lesions resolved by filling with bone from marginal sclerosis and reparative ossification centres. In the distal scapula and distal fibula, perichondrial new bone formation occurred that led to permanent enlargement of physeal regions. Angular limb deformity was not identified in any pig. CONCLUSIONS: Physeal osteochondrosis followed a similar dynamic development pattern to articular osteochondrosis. There were peaks in the incidence curves, compatible with failure of vessels during incorporation into bone. In some physes, osteochondrosis led to permanent enlargement, potentially relevant for decubital ulcers. The relationship between physeal osteochondrosis and angular limb deformity must be examined further in pigs over 6 months old in future.

Osteochondrosis in the Distal Femoral Physis of Pigs Starts With Vascular Failure.

Articular osteochondrosis (OC) arises due to vascular failure and ischemic chondronecrosis. The aim of the study was to describe the histological and computed tomographic (CT) characteristics of changes in the distal femoral physis of pigs, to determine if they represented OC lesions and if the pathogenesis was the same as for articular OC. The material included 19 male Landrace pigs bred for predisposition to OC. One or 2 pigs were euthanized and CT-scanned at 2-week intervals from 82 to 180 days of age. Material from 10 pigs was available for histological validation. The CT scans revealed 31 lesions confirmed in 3 planes and 1 additional macroscopically visible lesion confirmed in 2 CT planes. Twelve of the lesions were histologically validated. All lesions were compatible with OC. Cartilage canal and eosinophilic streak morphological changes corresponded to failure of end arteries coursing from the epiphysis, toward the metaphysis. The location of lesions was compatible with failure at the point of vessel incorporation into bone. Vascular failure was associated with retention of viable hypertrophic chondrocytes and delayed ossification but not cartilage necrosis. Lesion width ranged from 1.1% to 45.6% of the physis. Several lesions were expected to resolve due to small size and evidence of CT-identifiable, reparative ossification. Angular limb deformity was not detected in any pig. The pathogenesis of physeal OC started with vascular failure that was morphologically identical to articular OC. The heritable predisposition may therefore be the same. The association between lesions and limb deformity should be studied further in older pigs in future.

A QTL for Number of Teats Shows Breed Specific Effects on Number of Vertebrae in Pigs: Bridging the Gap Between Molecular and Quantitative Genetics.

Modern breeding schemes for livestock species accumulate a large amount of genotype and phenotype data which can be used for genome-wide association studies (GWAS). Many chromosomal regions harboring effects on quantitative traits have been reported from these studies, but the underlying causative mutations remain mostly undetected. In this study, we combine large genotype and phenotype data available from a commercial pig breeding scheme for three different breeds (Duroc, Landrace, and Large White) to pinpoint functional variation for a region on porcine chromosome 7 affecting number of teats (NTE). Our results show that refining trait definition by counting number of vertebrae (NVE) and ribs (RIB) helps to reduce noise from other genetic variation and increases heritability from 0.28 up to 0.62 NVE and 0.78 RIB in Duroc. However, in Landrace, the effect of the same QTL on NTE mainly affects NVE and not RIB, which is reflected in reduced heritability for RIB (0.24) compared to NVE (0.59). Further, differences in allele frequencies and accuracy of rib counting influence genetic parameters. Correction for the top SNP does not detect any other QTL effect on NTE, NVE, or RIB in Landrace or Duroc. At the molecular level, haplotypes derived from 660K SNP data detects a core haplotype of seven SNPs in Duroc. Sequence analysis of 16 Duroc animals shows that two functional mutations of the Vertnin (VRTN) gene known to increase number of thoracic vertebrae (ribs) reside on this haplotype. In Landrace, the linkage disequilibrium (LD) extends over a region of more than 3 Mb also containing both VRTN mutations. Here, other modifying loci are expected to cause the breed-specific effect. Additional variants found on the wildtype haplotype surrounding the VRTN region in all sequenced Landrace animals point toward breed specific differences which are expected to be present also across the whole genome. This Landrace specific haplotype contains two missense mutations in the ABCD4 gene, one of which is expected to have a negative effect on the protein function. Together, the integration of largescale genotype, phenotype and sequence data shows exemplarily how population parameters are influenced by underlying variation at the molecular level.

The use of deep learning to automate the segmentation of the skeleton from CT volumes of pigs.

Computed tomography (CT) scanning of pigs has been shown to produce detailed phenotypes useful in pig breeding. Due to the large number of individuals scanned and corresponding large data sets, there is a need for automatic tools for analysis of these data sets. In this paper, the feasibility of deep learning for fully automatic segmentation of the skeleton of pigs from CT volumes is explored. To maximize performance, given the training data available, a series of problem simplifications are applied. The deep-learning approach can replace our currently used semiautomatic solution, with increased robustness and little or no need for manual control. Accuracy was highly affected by training data, and expanding the training set can further increase performance making this approach especially promising.

Consequences of the natural course of articular osteochondrosis in pigs for the suitability of computed tomography as a screening tool.

BACKGROUND: A significant heritability has been documented for articular osteochondrosis. Selection against osteochondrosis has historically been based on macroscopic evaluation, but as computed tomography (CT) now is used to select boars with optimal body composition it can potentially also be used to screen for osteochondrosis. False negative diagnosis will occur if defects have not developed or have resolved prior to screening at a single time point. The aim of the current study was to assess the suitability of the use of CT at a single point in time as a screening tool in piglets for articular osteochondrosis, which is known to be a highly dynamic condition in which lesions develop and resolve over time. METHODS: Male Landrace piglets (n = 18) were serial CT scanned from 2-8 times at biweekly intervals from 70-180 days of age. At each interval, 1-2 piglets were euthanased and the left distal femur processed for histological validation. RESULTS: A total of 795 defects were identified in the 112 available CT scans. Within the hind and fore limbs, the incidence of defects was highest in the stifle (n = 321) and elbow joints (n = 110), respectively. Ninety-eight per cent of the defects in the stifle and elbow joints had developed by the 7th examination interval when the piglets were a mean age of 159 days old. The proportion of defects that resolved was lowest in the stifle joint at 51% and highest in the elbow joint at 69%. CONCLUSIONS: Scanning of the current piglets at an age of 159 days resulted in detection of 98% of the total number of defects that developed up to the maximum age of 180 days. The proportion of defects that resolved ranged from 51-69% for different joints, but may not adversely affect prevalence as this category of false negative diagnosis will result in selection of pigs that are disposed for healing. Optimally timed CT is a powerful screening tool for osteochondrosis.

Ossification defects detected in CT scans represent early osteochondrosis in the distal femur of piglets.

The purpose of the current study was to validate the use of CT for selection against osteochondrosis in pigs by calculating positive predictive value and comparing it to the positive predictive value of macroscopic evaluation, using histological examination as the reference standard. Eighteen male, hereditarily osteochondrosis-predisposed piglets underwent terminal examination at biweekly intervals from the ages of 82-180 days old, including CT scanning, macroscopic, and histological evaluation of the left distal femur. Areas of ischemic chondronecrosis (osteochondrosis) were confirmed in histological sections from 44/56 macroscopically suspected lesions, resulting in a positive predictive value of 79% (95% CI: 67-84%). Suspected lesions, that is; focal, radiolucent defects in the ossification front in CT scans corresponded to areas of ischemic chondronecrosis in 36/36 histologically examined lesions, resulting in a positive predictive value of 100% (95% CI: 90-100%). CT was superior to macroscopic evaluation for diagnosis of early stages of osteochondrosis in the distal femur of piglets. The current histologically validated observations can potentially be extrapolated to diagnostic monitoring of juvenile osteochondritis dissecans in children, or to animal models of human juvenile articular cartilage injury and repair.