Wageningen UR Livestock Research, PO Box 338, 6700 AH, Wageningen, the Netherlands.
Division Measure, Model and Manage Bioresponses, KU Leuven, PO Box 2456, 3001 Heverlee, Belgium.
J Dairy Sci. 2018 Jul;101(7):6322-6335. doi: 10.3168/jds.2017-13768. Epub 2018 Apr 5.
The objective of this study was to determine if a 3-dimensional computer vision automatic locomotion scoring (3D-ALS) method was able to outperform human observers for classifying cows as lame or nonlame and for detecting cows affected and nonaffected by specific type(s) of hoof lesion. Data collection was carried out in 2 experimental sessions (5 mo apart). In every session all cows were assessed for (1) locomotion by 2 observers (Obs1 and Obs2) and by a 3D-ALS; and (2) identification of different types of hoof lesions during hoof trimming (i.e., skin and horn lesions and combinations of skin/horn lesions and skin/hyperplasia). Performances of observers and 3D-ALS for classifying cows as lame or nonlame and for detecting cows affected or nonaffected by types of lesion were estimated using the percentage of agreement (PA), kappa coefficient (κ), sensitivity (SEN), and specificity (SPE). Observers and 3D-ALS showed similar SEN values for classifying lame cows as lame (SEN comparison Obs1-Obs2 = 74.2%; comparison observers-3D-ALS = 73.9-71.8%). Specificity values for classifying nonlame cows as nonlame were lower for 3D-ALS when compared with observers (SPE comparison Obs1-Obs2 = 88.5%; comparison observers-3D-ALS = 65.3-67.8%). Accordingly, overall performance of 3D-ALS for classifying cows as lame and nonlame was lower than observers (Obs1-Obs2 comparison PA = 84.2% and κ = 0.63; observers-3D-ALS comparisons PA = 67.7-69.2% and κ = 0.33-0.36). Similarly, observers and 3D-ALS had comparable and moderate SEN values for detecting horn (SEN Obs1 = 68.6%; Obs2 = 71.4%; 3D-ALS = 75.0%) and combinations of skin/horn lesions (SEN Obs1 = 51.1%; Obs2 = 64.5%; 3D-ALS = 53.3%). The SPE values for detecting cows without lesions when classified as nonlame were lower for 3D-ALS than for observers (SPE Obs1 = 83.9%; Obs2 = 80.2%; 3D-ALS = 60.2%). This was translated into a poor overall performance of 3D-ALS for detecting cows affected and nonaffected by horn lesions (PA Obs1 = 80.6%; Obs2 = 78.3%; 3D-ALS = 63.5% and κ Obs1 = 0.48; Obs2 = 0.44; 3D-ALS = 0.25) and skin/horn lesions (PA Obs1 = 75.1%; Obs2 = 75.9%; 3D-ALS = 58.6% and κ Obs1 = 0.35; Obs2 = 0.42; 3D-ALS = 0.10), when compared with observers. Performance of observers and 3D-ALS for detecting skin lesions was poor (SEN for Obs1, Obs2, and 3D-ALS <40%). Comparable SEN and SEN values for observers and 3D-ALS are explained by an overestimation of lameness by 3D-ALS when compared with observers. Thus, comparable SEN and SEN were reached at the expense high number of false positives and low SPE and SPE. Considering that observers and 3D-ALS showed similar performance for classifying cows as lame and for detecting horn and combinations of skin/horn lesions, the 3D-ALS could be a useful tool for supporting dairy farmers in their hoof health management.
本研究旨在确定三维计算机视觉自动运动评分（3D-ALS）方法是否能够优于人类观察者，用于对奶牛跛行或非跛行进行分类，以及检测特定类型蹄病变的奶牛。数据采集分两个实验阶段（相隔 5 个月）进行。在每个阶段，所有奶牛都由两名观察者（Obs1 和 Obs2）和 3D-ALS 进行（1）运动评估；并在蹄修整期间（即皮肤和角病变以及皮肤/角病变和皮肤/增生的组合）进行不同类型蹄病变的识别。使用一致性百分比（PA）、kappa 系数（κ）、灵敏度（SEN）和特异性（SPE）来评估观察者和 3D-ALS 对奶牛跛行或非跛行的分类以及对病变类型的奶牛的检测性能。观察者和 3D-ALS 对跛行奶牛的跛行分类的 SEN 值相似（SEN 比较 Obs1-Obs2 = 74.2%；比较观察者-3D-ALS = 73.9-71.8%）。与观察者相比，3D-ALS 对非跛行奶牛的非跛行分类的特异性值较低（SPE 比较 Obs1-Obs2 = 88.5%；比较观察者-3D-ALS = 65.3-67.8%）。因此，3D-ALS 对奶牛跛行和非跛行的分类总体性能低于观察者（Obs1-Obs2 比较 PA = 84.2%和 κ = 0.63；观察者-3D-ALS 比较 PA = 67.7-69.2%和 κ = 0.33-0.36）。同样，观察者和 3D-ALS 对 Horn 和皮肤/角病变的检测具有类似的、中等的 SEN 值（SEN Obs1 = 68.6%；Obs2 = 71.4%；3D-ALS = 75.0%）和皮肤/角病变的组合（SEN Obs1 = 51.1%；Obs2 = 64.5%；3D-ALS = 53.3%）。当被分类为无病变的奶牛时，3D-ALS 的 SPE 值低于观察者（SPE Obs1 = 83.9%；Obs2 = 80.2%；3D-ALS = 60.2%）。这导致 3D-ALS 对 Horn 病变和皮肤/角病变的奶牛的检测性能较差（PA Obs1 = 80.6%；Obs2 = 78.3%；3D-ALS = 63.5%和 κ Obs1 = 0.48；Obs2 = 0.44；3D-ALS = 0.25）和皮肤/角病变（PA Obs1 = 75.1%；Obs2 = 75.9%；3D-ALS = 58.6%和 κ Obs1 = 0.35；Obs2 = 0.42；3D-ALS = 0.10），与观察者相比。观察者和 3D-ALS 对皮肤病变的检测性能较差（Obs1、Obs2 和 3D-ALS 的 SEN <40%）。观察者和 3D-ALS 的可比 SEN 和 SEN 值解释为 3D-ALS 对跛行的高估与观察者相比。因此，在达到可比的 SEN 和 SEN 值的同时，假阳性的数量较高，特异性和特异性较低。考虑到观察者和 3D-ALS 对奶牛跛行和 Horn 病变以及皮肤/角病变的组合的分类性能相似，3D-ALS 可以成为支持奶牛场主进行蹄健康管理的有用工具。
