Automated Classification of Circulating Tumor Cells and the Impact of Interobsever Variability on Classifier Training and Performance

<div>The agreement, <svg height="14.1238pt" id="M21" style="vertical-align:-5.487901pt" version="1.1" viewbox="-0.0498162 -8.6359 21.4731 14.1238" width="21.4731pt" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink"><g transform="matrix(.013,0,0,-0.013,0,0)"><path d="M600 480C600 590 528 650 384 650H143L137 622C222 614 225 607 210 531L130 127C113 41 106 36 23 28L17 0H294L300 28C204 36 195 42 212 127L243 284L314 263C327 263 339 263 352 264C465 271 600 337 600 480ZM508 481C508 351 402 304 329 304C289 304 265 311 250 317L295 559C302 594 310 606 323 611C335 616 350 619 367 619C455 619 508 573 508 481Z" id="g113-81"></path><glyph.data ascent="3473" descent="-2876" horiz-adv-x="617" vert-adv-y="617"></glyph.data></g><g transform="matrix(.0091,0,0,-0.0091,6.656,3.132)"><path d="M682 0V33C615 38 603 49 571 134C507 305 441 494 378 665L343 655L136 136C101 48 86 41 22 33V0H246V33C170 41 161 50 183 112C196 152 209 192 224 232H441C461 176 478 126 493 87C504 49 499 41 433 33V0H682ZM425 279H242C273 363 303 448 335 533H337L425 279Z" id="g58-63"></path><glyph.data ascent="1007" descent="-369" horiz-adv-x="703" vert-adv-y="703"></glyph.data></g><g transform="matrix(.0091,0,0,-0.0091,12.925,3.132)"><path d="M472 439C432 432 382 428 334 425C303 442 271 451 236 451H235C158 451 54 398 54 285C54 216 97 171 140 152C124 133 93 105 53 90C52 81 55 63 64 48C77 27 102 5 137 -6C115 -22 77 -53 57 -73C40 -88 32 -107 32 -130C32 -189 96 -253 208 -253C340 -253 461 -159 461 -56C461 44 378 63 314 63C281 63 246 62 207 62C163 62 145 80 145 99C145 114 160 130 175 141C191 138 210 136 226 136C314 136 405 188 405 294C405 329 393 359 376 382L433 378C448 387 469 415 477 430L472 439ZM226 417C282 417 320 364 320 288S281 172 236 171C180 171 142 224 142 300S182 417 226 417ZM248 -9C291 -9 319 -11 346 -22S390 -57 390 -90C390 -151 339 -199 246 -199C171 -199 114 -159 114 -106C114 -79 132 -53 159 -30C178 -15 200 -9 248 -9Z" id="g58-101"></path><glyph.data ascent="1007" descent="-369" horiz-adv-x="479" vert-adv-y="479"></glyph.data></g><g transform="matrix(.0091,0,0,-0.0091,17.339,3.132)"><path d="M185 344V453C137 433 92 421 41 413V386C100 378 104 375 104 310V106C104 41 97 35 33 30V0H268V30C190 35 185 41 185 107V283C206 340 240 371 267 371C282 371 296 364 310 351C316 344 325 343 336 348C357 357 369 378 369 399C369 424 345 451 310 451C262 451 219 397 187 344H185Z" id="g58-112"></path><glyph.data ascent="1007" descent="-369" horiz-adv-x="380" vert-adv-y="380"></glyph.data></g></svg>, between observers across all <svg height="8.8423pt" id="M22" style="vertical-align:-0.2064009pt" version="1.1" viewbox="-0.0498162 -8.6359 44.761 8.8423" width="44.761pt" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink"><g transform="matrix(.013,0,0,-0.013,0,0)"><path d="M822 650H589L583 622C660 617 677 607 674 561C672 534 664 481 647 390L600 137H596L273 650H126L120 622C176 620 194 615 207 594C221 571 225 557 214 504L161 257C141 166 129 112 121 85C108 42 83 30 29 28L23 0H260L266 28C193 33 173 42 176 89C178 122 186 172 202 255L256 527H259L583 -8H612L690 390C708 481 720 535 728 558C744 603 756 619 816 622L822 650Z" id="g113-79"></path><glyph.data ascent="3473" descent="-2876" horiz-adv-x="839" vert-adv-y="839"></glyph.data></g><g transform="matrix(.013,0,0,-0.013,14.542,0)"><path d="M535 323V373H52V323H535ZM535 138V188H52V138H535Z" id="g117-34"></path><glyph.data ascent="3473" descent="-2876" horiz-adv-x="587" vert-adv-y="587"></glyph.data></g><g transform="matrix(.013,0,0,-0.013,25.805,0)"><path d="M137 343C167 482 260 545 321 574C357 591 397 603 429 609L423 641C382 634 335 622 295 608C189 570 37 457 37 238C37 84 125 -12 242 -12C362 -12 447 89 447 209C447 311 374 393 267 393C247 393 226 386 204 376L137 343ZM227 337C318 337 361 256 361 173C361 105 336 22 258 22C176 22 126 120 126 240C126 266 127 291 132 310C155 323 189 337 227 337Z" id="g113-55"></path><glyph.data ascent="3473" descent="-2876" horiz-adv-x="480" vert-adv-y="480"></glyph.data></g><g transform="matrix(.013,0,0,-0.013,32.045,0)"><path d="M384 0V27C293 34 287 42 287 114V635C232 613 172 594 109 583V559L157 557C201 555 205 550 205 499V114C205 42 199 34 109 27V0H384Z" id="g113-50"></path><glyph.data ascent="3473" descent="-2876" horiz-adv-x="480" vert-adv-y="480"></glyph.data></g><g transform="matrix(.013,0,0,-0.013,38.285,0)"><path d="M447 623H65C61 580 56 530 47 475H76C100 541 106 550 172 550H388C308 376 196 170 91 -1L98 -12L172 -2C268 204 360 408 455 611L447 623Z" id="g113-56"></path><glyph.data ascent="3473" descent="-2876" horiz-adv-x="480" vert-adv-y="480"></glyph.data></g></svg> cutouts.</div>

Journal of Immunology Research

fig2

Figure 2

Figure 2: Automated Classification of Circulating Tumor Cells and the Impact of Interobsever Variability on Classifier Training and Performance 