Citation: Long, X.; Ma, C.; Sheng, H.;
Chen, L.; Fei, Y.; Mi, L.; Han, D.; Ma,
J. Transfer Learning Analysis for
Subvisible Particle Flow Imaging of
Pharmaceutical Formulations. Appl.
Sci. 2022, 12, 5843. https://doi.org/
10.3390/app12125843
Academic Editors: Keun Ho Ryu and
Nipon Theera-Umpon
Received: 29 April 2022
Accepted: 7 June 2022
Published: 8 June 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Article
Transfer Learning Analysis for Subvisible Particle Flow
Imaging of Pharmaceutical Formulations
Xiangan Long
1
, Chongjun Ma
2
, Han Sheng
1
, Liwen Chen
3,4
, Yiyan Fei
3
, Lan Mi
3
, Dongmei Han
2,
*
and Jiong Ma
1,3,5,
*
1
Institute of Biomedical Engineering and Technology, Academy for Engineering and Technology,
Fudan University, 220 Handan Road, Shanghai 200433, China; 19210860049@fudan.edu.cn (X.L.);
19110860077@fudan.edu.cn (H.S.)
2
Process Development Downstream & Formulation, Shanghai Henlius Biotech, Inc., 1801 Hongmei Road,
Shanghai 200233, China; chongjun_ma@henlius.com
3
Shanghai Engineering Research Center of Ultra-Precision Optical Manufacturing, Key Laboratory of Micro
and Nano Photonic Structures (Ministry of Education), Green Photoelectron Platform, Department of Optical
Science and Engineering, Fudan University, 220 Handan Road, Shanghai 200433, China;
13681741630@163.com (L.C.); fyy@fudan.edu.cn (Y.F.); lanmi@fudan.edu.cn (L.M.)
4
Ruidge Biotech Co., Ltd., Lin-Gang Special Area, China (Shanghai) Pilot Free Trade Zone, No. 888,
Huanhu West 2nd Road, Shanghai 200131, China
5
Shanghai Engineering Research Center of Industrial Microorganisms, The Multiscale Research Institute of
Complex Systems (MRICS), School of Life Sciences, Fudan University, 220 Handan Road,
Shanghai 200433, China
* Correspondence: dongmei_han@henlius.com (D.H.); jiongma@fudan.edu.cn (J.M.)
Abstract:
Subvisible particles are an ongoing problem in biotherapeutic injectable pharmaceutical
formulations, and their identification is an important prerequisite for tracing them back to their
source and optimizing the process. Flow imaging microscopy (FIM) is a favored imaging technique,
mainly because of its ability to achieve rapid batch imaging of subvisible particles in solution with
excellent imaging quality. This study used VGG16 after transfer learning to identify subvisible
particle images acquired using FlowCam. We manually prepared standards for seven classes of
particles, acquired the image information through FlowCam, and fed the images over 5
µ
m into
VGG16 consisting of a convolutional base of VGG16 pre-trained with ImageNet data and a custom
classifier for training. An accuracy of 97.51% was obtained for the test set data. The study also
demonstrated that the recognition method using transfer learning outperforms machine learning
methods based on morphological parameters in terms of accuracy, and has a significant training
speed advantage over scratch-trained CNN. The combination of transfer learning and FIM images is
expected to provide a general and accurate data-analysis method for identifying subvisible particles.
Keywords:
image analysis; protein formulation; convolutional neural network; particle identification;
machine learning
1. Introduction
Therapeutic protein formulations are among the fastest-developing pharmaceutical
classes. Pharmaceutical formulations always contain subvisible particles owing to the
limitations of the production process and the instability of the proteins [
1
]. According to
the discussion in USP1790 [
2
], particles in protein formulations can usually be classified as
“extrinsic”, “intrinsic”, or “inherent” particles. Extrinsic particles can be considered foreign
to the manufacturing process, including hair, non-process-related fibers, starch, minerals,
and similar inorganic and organic materials. Intrinsic particles are those related to the
manufacturing process, which may originate from processing equipment or packaging
materials, including seals, gaskets, packaging glass and elastomers, fluid transport tubing,
and silicone lubricant. Inherent particles are associated with the production formulations
Appl. Sci. 2022, 12, 5843. https://doi.org/10.3390/app12125843 https://www.mdpi.com/journal/applsci
